Factors associated with Silent Cerebral Microbleeds in Hemodialysis Patients
Toshihide Naganumaa, Yoshiaki Takemotoa, Takeshi Yamasakia, Hideaki Shimab, Tetsuo Shojib,
Eiji Ishimurab, Yoshiki Nishizawab, Michiharu Morinoc, Mikio Okamurad, Tatsuya Nakatania
a
Department of Urology, bDepartment of Nephrology, cDepartment of Neurosurgery, Osaka City
University Graduate School of Medicine, Osaka, Japan.
d
Department of Nephrology, Ohno Memorial Hospital, Osaka, Japan
Correspondence to:
Toshihide Naganuma, MD
Department of Urology, Osaka City University Medical School
1-4-3 Asahi-machi, Abeno-ku, Osaka, 545-8585, Japan
Phone: +81-6-6645-3857
Fax: +81-6-6647-4426
E-mail: spxd48k9@aria.ocn.ne.jp
Running title: Silent Cerebral Microbleeds in HD Patients
1
ABSTRACT
Background.
The recent development of gradient-echo T2*-weighted magnetic resonance
imaging (MRI) has enabled the highly accurate detection of prior cerebral microbleeds (CMBs),
which might indicate a higher risk of future intracerebral hemorrhage (ICH) and be a marker of
cerebral small-vessel disease in the general population.
The present study investigated the clinical
factors associated with the presence of CMBs in hemodialysis (HD) patients.
Methods.
Cranial MRI, including T2*-weighted MRI, was performed on 179 HD patients without
symptomatic cerebrovascular disease and 58 healthy control subjects, and we investigated the
prevalence of CMBs and clinical associated factors with the presence of CMBs.
We also
investigated the relationship between CMBs and other cerebral small-vessel diseases.
Results.
The prevalence of CMBs was significantly higher in HD patients than in the healthy
subjects {45 patients (25.1%) vs. none in the healthy controls (0%), P  0.0001}.
Multiple logistic
regression analysis showed that independent and significant factors associated with the presence of
CMBs were age, systolic blood pressure and diastolic blood pressure.
Moreover, the presence of
CMBs correlated significantly with the presence of lacunar infarcts, periventricular hyperintensity
and deep and subcortical white matter hyperintensity.
Conclusions.
These findings indicated a high prevalence of CMBs among HD patients, and that
older age and high blood pressure were strong factors associated with the presence of CMBs.
Moreover, CMBs were closely associated with other cerebral small-vessel diseases.
Key words: magnetic resonance imaging; T2*-weighted MRI; cerebral microbleeds; lacunar
infarcts; periventricular hyperintensity; subcortical white matter hyperintensity; hemodialysis
2
INTRODUCTION
The highly accurate detection of prior cerebral microbleeds (CMBs), that are hardly
detectable by conventional CT scanning and magnetic resonance imaging (MRI), has been made
possible with the recent development of gradient-echo T2*-weighted MRI [1].
CMBs are seen as
small, round hypointense lesions on T2*-weighted MRI, pathologically representing focal
hemosiderin deposition associated with previous microhemorrhages resulting from lipohyalinized
cerebral arterioles, and are related to bleeding-prone small-vessel diseases, such as hypertensive
vasculopathy (arteriolosclerotic microangiopathy) and cerebral amyloid angiopathy (CAA) [1-4].
The prevalence rates of CMBs in healthy populations without cerebrovascular disease range from
3.1% to 8.5% [5-6].
In contrast, the prevalence of CMBs has been shown to be nearly 10-fold
greater in cohorts with spontaneous intracerebral hemorrhage (ICH) (range 47% to 80%) compared
with the healthy elderly [5-6].
In addition, the distribution of CMBs has been reported to
correlate with the location of ICH [3].
Because ICH has been thought to occur as a result of
small-artery disease of the brain, CMBs have been emphasized as a risk factor for subsequent stroke
attack, particularly for symptomatic ICH, and might be a marker of cerebral small-vessel disease
[5-6].
In hemodialysis (HD) patients, cerebrovascular disease is one of the major causes of death
[7]. It has been reported that HD patients have a much higher incidence of stroke than normal
populations [7-8].
Iseki et al. [7] reported that the incidences of ICH and cerebral infarction were
8.7 and 2.2 per 1000 person-years, respectively, and that the relative risk compared to the general
population in Okinawa was 10.7 for ICH and 2.0 for cerebral infarction in HD patients.
In other
words, strokes in HD patients are characterized by a higher incidence of hypertensive ICH
compared with the general population [7].
For the prevention of ICH among HD patients, it would
be helpful to identify patients with cerebral small-vessel changes, as evidenced by CMBs, as well as
to identify the risk factors for CMBs.
However, there have been only two reports [9-10]
describing CMBs in HD patients, and the clinical significance of CMBs in HD patients has not yet
3
been elucidated.
In the present study, we investigated the clinical factors associated with the
presence of CMBs and the relationship between CMBs and other cerebral small-vessel diseases in
HD patients.
4
METHODS
Patients
We performed a cross-sectional study of 179 neurologically-normal HD patients in stable
condition who agreed to undergo MRI from April 2006 to October 2009 at Osaka City University
Hospital or Ohno Memorial Hospital. This study protocol was approved by the ethical committee
of Osaka City University School of Medicine (No.1415).
subjects prior to their participation in the study.
Informed consent was obtained from all
The subjects were selected from the total
population of hemodialysis patients (n=206) treated at Osaka City University Hospital or Ohno
Memorial Hospital. Out of the 206 hemodialysis patients, 21 were excluded because of a previous
history or symptoms of stroke, autosomal dominant polycystic kidney disease, chronic infection,
chronic inflammatory disease or malignant disease. Six patients did not agree to participate. The
remaining 187 patients were studied.
As a control group, 58 age- and sex-matched healthy subjects who visited Ohno Memorial
Hospital for screening check-up of the brain were also examined.
All control subjects were
enrolled consecutively after we presented and explained the study protocol.
None of the healthy
control subjects had chronic kidney disease (CKD).
Magnetic resonance imaging
All participating patients had a brain MRI that used a superconducting magnet at a field
strength of 1.5 T on proton density, T1-, T2- weighted FLAIR MRI and two-dimensional gradient
echo (repetition time 606 ms, echo time 23 ms, flip angle 18°- 20°, field of view 200 mm, acquisition
matrix 256 x 175- 224) T2*- weighted MRI in axial planes at 5 mm thick slices with an interslice gap
of 1.5 mm.
CMBs were defined as focal areas of signal loss consisting of homogeneous, rounded
lesions with diameters of 2 to 5 mm on T2*-weighted MRI.
Hypointensities of the globus pallidus,
which probably represented calcification or flow void artifacts of the pial vessel, were excluded.
Lacunar infarcts were defined as focal areas ≥ 3 mm and < 15 mm in diameter on both T15
and T2-weighted MRI, which were visible as low-signal intensity areas on T1-weighted MRI and as
high signal intensity areas on T2-weighted MRI.
White matter hyperintensities, such as periventricular hyperintensity (PVH) and deep and
subcortical white matter hyperintensity (DSWMH), were defined as focal areas of increased signal
intensity on proton density-weighted and T2-weighted MRI, if there was no corresponding signal
alteration on T1-weighted MRI.
The MR images were assessed independently by two neuroradiologists who were blinded
to any clinical information.
As cases with inconsistencies were excluded from the study, out of
187 patients that underwent MRI, 8 cases with diagnostic inconsistencies were excluded, resulting
in 179 cases that were examined for this study.
Risk factors
To evaluate the clinical factors associated with the presence of CMBs, we investigated age,
gender, dialysis duration, blood pressure, interdialytic body weight gain, serum albumin, C-reactive
protein, hemoglobin, the presence or absence of hypertension, diabetes mellitus (DM), dyslipidemia,
ischemic heart disease or current smoking, and the use of antithrombotic (antiplatelet/anticoagulant)
or antihypertensive agents.
Hypertension was defined by (a) the administration of antihypertensive agents and/or a
history of this disorder; (b) a systolic blood pressure (SBP) greater than 140 mmHg; or (c) a
diastolic blood pressure (DBP) greater than 90 mmHg.
Blood pressure was determined during the
pre- and post-HD sessions.
DM was defined as a person using oral antidiabetic drugs or insulin, or when fasting blood
glucose was ≥ 126 mg/dL.
Dyslipidemia was defined as present if the subject had a total cholesterol > 220 mg/dl,
triglyceride > 150 mg/dl and high-density lipoprotein cholesterol < 40 mg/dl, or had received
medical treatment for hyperlipidemia.
Blood samples were taken from the arterial line before the HD sessions.
6
Ischemic heart disease was defined as either angina or a history of myocardial infarction,
coronary artery bypass surgery or percutaneous transluminal angioplasty.
Angina was diagnosed
by exercise electrocardiography, myocardial perfusion imaging or coronary angiography.
Statistical analysis
Data were summarized as a percentage or as the mean  standard deviation (SD), whenever
appropriate.
Differences between groups were examined with Student’s t-test. Categorical
variables were compared using chi-squared analysis.
Logistic regression analysis was used to
assess the influence of variables on the presence of CMBs.
To determine the association between
hypertension and the presence of CMBs, we used several methods to define hypertension in logistic
regression analysis.
Models 1 through 6 included pre-dialysis SBP, pre-dialysis DBP, the presence
of hypertension, post-dialysis SBP, post-dialysis DBP and the use of antihypertensive agents,
respectively, as variables for hypertension.
Gender, the presence or absence of CMBs, lacunar
infarcts, PVH, DSWMH, DM, dyslipidemia, ischemic heart disease or current smoking, use of
antithrombotic agents and antihypertensive agents were represented by dummy variables 1 = male,
0 = female; 1 = presence, 0 = absence; smoker = 1, non- smoker =0; yes = 1, no =0 in logistic
analysis.
A P-value of  0.05 was considered statistically significant.
These results were
obtained using Stat View V Statistical Software (SAS Institute Inc., Cary, NC, USA).
7
RESULTS
Characteristics of patients with and without CMBs, and the healthy subjects (Table 1)
Table 1 presents the characteristics of the 179 HD patients and 58 healthy subjects.
The
prevalence of CMBs was significantly higher in the HD patients than in the healthy subjects {45
patients (25.1%) vs. none in the healthy controls (0%), P  0.0001}.
The patients with CMBs were significantly older than those without CMBs.
SBP (pre-
and post-dialysis) and the prevalence of hypertension and DM in the patients with CMBs were
significantly higher than those in the patients without CMBs.
There was no significant difference
in the other factors between the patients with/without CMBs.
Twenty-eight patients (62.2%) had multiple CMBs with counts ranging from 2 to 22.
Among the 45 patients with CMBs, 20 (44.4%) were located in the subcortical white matter, 15
(33.7%) in the thalamus, 29 (64.4%) in the basal ganglia, 14 (31.1%) in the brain stem and 7
(15.6%) in the cerebellum.
Univariate logistic regression analysis of factors associated with CMBs (Table 2)
Univariate logistic regression analysis was performed to examine the factors associated
with the presence of CMBs (Table 2).
Independent and significant factors associated with the
presence of CMBs were age, SBP (pre- and post-dialysis) and the presence of hypertension and DM,
while DBP was at borderline significance.
Multiple logistic regression analysis of factors associated with CMBs (Table 3 and 4)
Multiple logistic regression analysis was performed to examine the factors associated with
the presence of CMBs independent of other factors (Tables 3 and 4). Models 1 through 6 included
pre-dialysis SBP, pre-dialysis DBP, presence of hypertension, post-dialysis SBP, post-dialysis DBP
and the use of antihypertensive agents, respectively, as variables for hypertension.
In these models,
age, SBP (pre- and post-dialysis) and DBP (pre- and post-dialysis) were indicated as independent
and significant factors associated with the presence of CMBs, whereas the presence of hypertension
and the use of antihypertensive agents were at borderline significance.
8
Prevalence of cerebral small-vessel diseases with and without CMBs (Table 5)
The prevalence of lacunar infarcts, PVH and DSWMH was 41.3% (74/179), 44.1%
(79/179) and 46.9% (84/179), respectively (Table 5).
The prevalence of lacunar infarcts, PVH and
DSWMH in the patients with CMBs was significantly higher than that in the patients without
CMBs.
Relationship between CMBs and other cerebral small-vessel diseases (Table 6)
Univariate logistic regression analysis was performed to examine the relationship between
CMBs and other cerebral small-vessel diseases (Table 6).
The presence of CMBs was
significantly associated with that of lacunar infarcts, PVH and DSWMH.
Furthermore, in logistic
regression analysis after adjustment for age and gender, which were unmodifiable basic clinical
factors of the patients, the positive association between CMBs and the presence of PVH and
DSWMH remained significant, whereas the presence of lacunar infarcts was at borderline
significance.
9
DISCUSSION
In the present study, we investigated 1) the prevalence of T2*-weighted MRI diagnosed
CMBs in HD patients and healthy subjects, 2) the clinical factors associated with the presence of
CMBs and 3) the association of CMBs with other cerebral small-vessel diseases in HD patients.
As a result, the prevalence of CMBs was significantly higher in HD patients than in the healthy
subjects. Multiple logistic regression analysis showed that independent and significant factors
associated with the presence of CMBs were age and blood pressure.
Moreover, the presence of
CMBs correlated with that of other cerebral small-vessel diseases, namely lacunar infarcts, PVH
and DSWMH.
In recent years, with the development of T2*-weighted MRI, small foci of chronic blood
products in brain tissues, known as CMBs, have been increasingly recognized in different
populations [5-6].
In HD patients, Yokoyama et al. [10] reported that the overall prevalence of
CMBs was 19.3% (11/57) in those HD patients without previous stroke, and the prevalence of
CMBs was significantly greater in HD patients compared with control patients without chronic
renal failure. Watanabe [9] reported that the overall prevalence of CMBs was 35% (28/80) in HD
patients (including patients with previous stroke).
In our study, the overall prevalence of CMBs in
HD patients without previous stroke was 25.1% (45/179) and was significantly higher compared to
that of the healthy subjects, which was consistent with these previous reports [9-10], indicating a
higher prevalence of CMBs in HD cohorts compared with healthy populations.
As the risk factors for CMBs, older age [5-6, 11-13], hypertension [4-6, 12], smoking habit
[14], male sex [15], DM [13], low total cholesterol [11, 16], retinal microvascular lesions [17] and
left ventricular hypertrophy [18] have been demonstrated in cohorts with strokes or
community-dwelling elderly individuals.
In particular, older age and hypertension are most
commonly associated with CMBs. Because older age, hypertension, DM and left ventricular
hypertrophy are common in HD patients, it is thought that they may have multiple risk factors for
CMBs, thus, explaining the high prevalence of CMBs in HD patients.
10
In the present study, logistic regression analysis was performed to examine the factors
associated with the presence of CMBs.
As a result, age, SBP and DBP were independent and
significant factors associated with the presence of CMBs; consistent with previous studies [5-6, 12]
in non-renal populations.
In general, the prevalence of CMBs increases with age [5-6, 11-13],
likely due to advanced age-associated small-vessel diseases, such as hypertensive vasculopathy and
CAA. Hypertension is the major cause of lipohyalinosis and microaneurysms [2], which might
cause CMBs in HD patients, as well as in cohorts with non-renal populations.
However,
Yokoyama et al. [9] reported that there was no significant difference between CMBs and the
incidence of hypertension in HD patients.
The discrepancy between the studies by Yokoyama [9]
and ours may derive from the size of the study and the analytical methods implemented.
Our
study evaluated a greater number of patients (179 vs. 57). Moreover, they did not perform logistic
regression analysis, and did not mention SBP and DBP.
In the present study, when evaluating the
association between hypertension and CMBs, we examined both continuous (SBP and DBP) and
categorical variables (the presence of hypertension and the use of antihypertensive agents), but no
association was found using the categorical variable, which was, in part, consistent with
Yokoyama’s report [9].
In the present study, we found no significant association between the presence of CMBs
and the duration of dialysis, which was consistent with previous reports [9-10].
Moreover, our
findings were consistent with the results of previous studies that atherosclerosis of major arteries
might not be accelerated by HD itself [19].
In general populations, arteriosclerosis and
cardiovascular disease occur during the advancement from the periods of chronic kidney disease
stages 2 to 3 [20], which suggests the CMBs might develop during the pre-dialysis period.
Further
examination of CMBs in pre-dialysis chronic kidney disease patients may resolve this issue.
It is well established that the use of antithrombotic drugs (antiplatelet/anticoagulant drugs)
is related to the increased incidence and recurrence of ICH [21]. Similarly, the development of
CMBs also may be accelerated by the use of these antithrombotic drugs.
11
Recently, Vernooji et al.
[22] reported that the use of platelet aggregation inhibitors was related to the presence of CMBs in a
population-based cross-sectional study.
By contrast, in the present study, there was no correlation
between the use of antithrombotic therapy and the presence of CMBs.
In almost all HD patients,
anticoagulation therapy (systemic heparinization) was necessary during repeated HD sessions, and
there were many occasions for the use of antithrombotic therapy, such as the treatment of coronary
heart disease, arteriosclerosis obliterans, peripheral arterial disease, ischemic cerebrovascular
disease and vascular access trouble. Thus, antithrombotic therapy would not likely be found as an
independent factor associated with the presence of CMBs in HD patients.
In addition, the high prevalence of CMBs has been observed with other cerebral
small-vessel diseases, including lacunar infarcts [3, 5-6], white matter disease (PVH, DSWMH) [3,
5-6], CAA [5-6] and cerebral autosomal dominant arteriopathy with subcortical infarcts and
leukoencephalopathy [23].
in the general population.
Therefore, CMBs might be a marker of cerebral small-vessel disease
In our present study in HD patients, lacunar infarcts and white matter
disease (PVH, DSWMH) were also found by MRI (Table 5), and we observed that the presence of
CMBs correlated significantly with that of lacunar infarcts and white matter disease (PVH,
DSWMH) (Table 6), showing the close association between CMBs and other cerebral small-vessel
diseases in HD patients.
These results indicated that CMBs might be as a marker for cerebral
small-vessel changes in HD patients as in general populations.
This study has several limitations. First, because of the cross-sectional design, the results
do not indicate causality, but this point will be further elucidated in longitudinal studies. Second,
as markers of underlying bleeding-prone small-vessel diseases, CMBs might predict the future risk
of symptomatic ICH in HD patients, as well as cohorts with strokes [24-25]. Thus, further
prospective studies are needed to elucidate the prognostic role of CMBs in HD patients.
Third,
although we reported the prevalence of CMBs based on conventional two-dimensional
T2*-weighted MRI, the newer techniques, such as three-dimensional T2*-weighted MRI, might
have provided different results [11-12].
Thus, further examination should be performed using
12
three-dimensional T2*-weighted MRI.
In conclusion, the present study demonstrated a high prevalence of CMBs in HD patients,
that older age and high blood pressure were independent and significant factors associated with the
presence of CMBs and that there was a close association between CMBs and other types of cerebral
small-vessel diseases.
Further longitudinal studies are needed to elucidate whether CMBs can
predict symptomatic ICH in HD patients, and how we should manage patients with CMBs to
prevent ICH.
13
Acknowledgements
This work was supported by grants from the Osaka City University Medical Research Foundation.
14
References
1.
Offenbacher H, Fazekas F, Schmidt R, Koch M, Fazekas G, Kapeller P. MR of cerebral
abnormalities concomitant with primary intracerebral hematomas. AJNR Am J Neuroradiol. 1996;
17:573-8.
2.
Fazekas F, Kleinert R, Roob G, Kleinert G, Kapeller P, Schmidt R, Hartung HP . Histopathologic
analysis of foci of signal loss on gradient-echo T2*-weighted MR images in patients with spontaneous
intracerebral hemorrhage: evidence of microangiopathy-related microbleeds. AJNR Am J Neuroradiol.
1999; 20:637-42.
3.
Kato H, Izumiyama M, Izumiyama K, Takahashi A, Itoyama Y. Silent cerebral microbleeds on
T2*-weighted MRI: correlation with stroke subtype, stroke recurrence, and leukoaraiosis. Stroke. 2002;
33:1536-40.
4.
Tanaka A, Ueno Y, Nakayama Y, Takano K, Takebayashi S. Small chronic hemorrhages and
ischemic lesions in association with spontaneous intracerebral hematomas. Stroke. 1999; 30:1637-42.
5.
Koennecke HC. Cerebral microbleeds on MRI: prevalence, associations, and potential clinical
implications. Neurology. 2006; 66:165-71.
6.
Viswanathan A, Chabriat H. Cerebral microhemorrhage. Stroke. 2006; 37:550-5.
7.
Iseki K, Kinjo K, Kimura Y, Osawa A, Fukiyama K. Evidence for high risk of cerebral
hemorrhage in chronic dialysis patients. Kidney Int. 1993; 44:1086-90.
8.
Seliger SL, Gillen DL, Longstreth WT, Jr., Kestenbaum B, Stehman-Breen CO. Elevated risk of
stroke among patients with end-stage renal disease. Kidney Int. 2003; 64:603-9.
9.
Watanabe A. Cerebral microbleeds and intracerebral hemorrhages in patients on maintenance
hemodialysis. J Stroke Cerebrovasc Dis. 2007; 16:30-3.
10.
Yokoyama S, Hirano H, Uomizu K, Kajiya Y, Tajitsu K, Kusumoto K . High incidence of
microbleeds in hemodialysis patients detected by T2*-weighted gradient-echo magnetic resonance
imaging. Neurol Med Chir (Tokyo). 2005; 45:556-60; discussion 60.
11.
Vernooij MW, van der Lugt A, Ikram MA, Wielopolski PA, Niessen WJ, Hofman A, Krestin GP,
Breteler MM. Prevalence and risk factors of cerebral microbleeds: the Rotterdam Scan Study. Neurology.
2008; 70:1208-14.
12.
Greenberg SM, Vernooij MW, Cordonnier C, Viswanathan A, Al-Shahi Salman R, Warach S,
Launer LJ, Van Buchem MA, Breteler MM. Cerebral microbleeds: a guide to detection and interpretation.
Lancet Neurol. 2009; 8:165-74.
13.
Nighoghossian N, Hermier M, Adeleine P, Blanc-Lasserre K, Derex L, Honnorat J, Philippeau F,
Dugor JF, Froment JC, Trouillas P. Old microbleeds are a potential risk factor for cerebral bleeding after
ischemic stroke: a gradient-echo T2*-weighted brain MRI study. Stroke. 2002; 33:735-42.
14.
Tsushima Y, Tanizaki Y, Aoki J, Endo K. MR detection of microhemorrhages in neurologically
healthy adults. Neuroradiology. 2002; 44:31-6.
15.
Jeerakathil T, Wolf PA, Beiser A, Hald JK, Au R, Kase CS, Massaro JM, DeCarli C. Cerebral
microbleeds: prevalence and associations with cardiovascular risk factors in the Framingham Study.
Stroke. 2004; 35:1831-5.
16.
Lee SH, Bae HJ, Yoon BW, Kim H, Kim DE, Roh JK. Low concentration of serum total
cholesterol is associated with multifocal signal loss lesions on gradient-echo magnetic resonance
imaging: analysis of risk factors for multifocal signal loss lesions. Stroke. 2002; 33:2845-9.
17.
Qiu C, Cotch MF, Sigurdsson S, Garcia M, Klein R, Jonasson F, Klein BE, Eiriksdottir G, Harris
TB, van Buchem MA, Gudnason V, Launer LJ. Retinal and cerebral microvascular signs and diabetes:
the age, gene/environment susceptibility-Reykjavik study. Diabetes. 2008; 57:1645-50.
18.
Lee SH, Park JM, Kwon SJ, Kim H, Kim YH, Roh JK, Yoon BW. Left ventricular hypertrophy is
associated with cerebral microbleeds in hypertensive patients. Neurology. 2004; 63:16-21.
19.
Shoji T, Emoto M, Tabata T, Kimoto E, Shinohara K, Maekawa K, Kawagishi T, Tahara H,
Ishimura E, Nishizawa Y. Advanced atherosclerosis in predialysis patients with chronic renal failure.
Kidney Int. 2002; 61:2187-92.
20.
Zhang L, Zhao F, Yang Y, Qi L, Zhang B, Wang F, Wang S, Liu L, Wang H. Association between
carotid artery intima-media thickness and early-stage CKD in a Chinese population. Am J Kidney Dis.
2007; 49:786-92.
21.
He J, Whelton PK, Vu B, Klag MJ. Aspirin and risk of hemorrhagic stroke: a meta-analysis of
15
randomized controlled trials. JAMA. 1998; 280:1930-5.
22.
Vernooij MW, Haag MD, van der Lugt A, Hofman A, Krestin GP, Stricker BH, Breteler MM. Use
of Antithrombotic Drugs and the Presence of Cerebral Microbleeds: The Rotterdam Scan Study. Arch
Neurol. 2009.
23.
Lesnik Oberstein SA, van den Boom R, van Buchem MA, van Houwelingen HC, Bakker E,
Vollebregt E, Ferrari MD, Breuning MH, Haan J. Cerebral microbleeds in CADASIL. Neurology. 2001;
57:1066-70.
24.
Boulanger JM, Coutts SB, Eliasziw M, Gagnon AJ, Simon JE, Subramaniam S, Sohn CH, Scott
J, Demchuk AM. Cerebral microhemorrhages predict new disabling or fatal strokes in patients with
acute ischemic stroke or transient ischemic attack. Stroke. 2006; 37:911-4.
25.
Soo YO, Yang SR, Lam WW, Wong A, Fan YH, Leung HH, Chan AY, Leung C, Leung TW, Wong
LK. Risk vs benefit of anti-thrombotic therapy in ischaemic stroke patients with cerebral microbleeds. J
Neurol. 2008; 255:1679-86.
16