frequency and location of microbleeds - roob

FazekasGudrun Roob, Anita Lechner, Reinhold Schmidt, Erich Flooh, Hans-Peter Hartung and Franz

HemorrhageFrequency and Location of Microbleeds in Patients With Primary Intracerebral

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright 2000 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/01.STR.31.11.2665

2000;31:2665-2669Stroke.

http://stroke.ahajournals.org/content/31/11/2665World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://stroke.ahajournals.org//subscriptions/is online at: Stroke Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer process is available in theRequest Permissions in the middle column of the Web page under Services. Further information about thisOnce the online version of the published article for which permission is being requested is located, click

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Strokein Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on September 15, 2014http://stroke.ahajournals.org/Downloaded from by guest on September 15, 2014http://stroke.ahajournals.org/Downloaded from

Frequency and Location of Microbleeds in Patients WithPrimary Intracerebral Hemorrhage

Gudrun Roob, MD; Anita Lechner, MD; Reinhold Schmidt, MD; Erich Flooh, MSc;Hans-Peter Hartung, MD; Franz Fazekas, MD

Background and PurposeMRI is known to detect clinically silent microbleeds (MBs) in patients with primaryintracerebral hemorrhage (pICH), but the frequency and diagnostic and clinical significance of this finding are stilldebated. Therefore, we investigated a consecutive series of pICH patients and analyzed the patterns of MB distributionin the context of clinical variables and location of the symptomatic hematoma.

MethodsThe study population consisted of 109 patients with pICH. There were 59 women and 50 men aged 22 to 91years (mean 64.6 years). MRI was obtained on a 1.5-T system with use of a gradient-echo T2*-weighted sequence. Acohort of 280 community-dwelling asymptomatic elderly individuals who underwent the same imaging protocol servedfor comparison.

ResultsMBs were seen in 59 (54%) patients and ranged in number from 1 to 90 lesions (mean 14, median 6). In themajority of patients, MBs were located simultaneously in various parts of the brain, with a preference forcortical-subcortical regions (39%) and the basal ganglia/thalami (38%). There was some tendency toward a regionalassociation between MB location and the site of the symptomatic hematoma, but we could not discern specific patternsof MB distribution. Logistic regression analysis identified MBs, periventricular hyperintensity grades, and lacunes butnot risk factors as independent variables contributing to a correct classification of pICH and control individuals.

ConclusionsMBs can be detected in more than half of the patients with pICH and appear to be quite general markersof various types of bleeding-prone microangiopathy. (Stroke. 2000;31:2665-2669.)Key Words: etiology n hemosiderin n intracerebral hemorrhage n magnetic resonance imaging n microcirculation

Magnetic resonance imaging (MRI) has the potential toreveal residues of intracerebral bleeding throughoutlife because of its high sensitivity for iron-containing com-pounds. At the site of intracerebral hemorrhage (ICH),hemosiderin remains stored in macrophages and leads tofocal dephasing of the MRI signal. This causes areas of pastbleeding to appear dark on T2-weighted images. Gradient-echo techniques with high sensitivity for differences inmagnetic susceptibility enhance these effects of hemosider-in.1 Previous studies have called attention to the frequentobservation of small areas of signal loss in patients withprimary ICH (pICH), which were suggested to representresidues of earlier clinically silent microbleeds (MBs).2,3Subsequently, this assumption was supported by correlativehistopathologic data, which confirmed perivascular hemosid-erin deposition around angiopathic arterioles as the predom-inant cause for such a finding.4,5 The association withsmall-vessel diseases explains that MBs were also reported inpatients with cerebral ischemic damage5,6 and even in a smallproportion of asymptomatic elderly individuals.7 However,the clinical and diagnostic significance of this finding is notyet fully understood.

In patients with pICH, the reported frequency of earlierMBs ranges from 17% to 80%.8 These large differences areprobably a consequence of small patient groups examined,differences in patient selection, and the inconsistent use ofeither conventional spin-echo or gradient-echo MR tech-niques or both.8 Although hypertensive microangiopathyappears to be their prevailing cause, MBs have also beenstrongly associated with the presence of cerebral amyloidangiopathy.9 In this context, a cortical-subcortical appearanceof MBs has been suggested to be an almost pathognomonicfinding.3 On the basis of this assumption, it could bespeculated that certain patterns in the distribution of MBsmay serve to indicate different types of microangiopathy. Theamount of predictive information contained in the detectionof MBs and possible concerns regarding the use of antico-agulants in such patients have also been debated.

We attempted to provide further information on theseaspects by performing a detailed analysis of the frequencyand distribution patterns of MBs in the context of clinicalvariables and hematoma location in a large consecutive seriesof patients with pICH. We also performed a comparison with

Received June 9, 2000; final revision received July 3, 2000; accepted July 12, 2000.From the Department of Neurology (G.R., A.L., R.S., E.F., H.-P.H., F.F.) and the MR Institute (A.L., R.S., F.F.), Karl-Franzens University, Graz,

Austria.Correspondence to Franz Fazekas, MD, Department of Neurology, Karl-Franzens University, Auenbruggerplatz 22, A-8036 Graz, Austria. E-mail

[email protected] 2000 American Heart Association, Inc.Stroke is available at http://www.strokeaha.org

2665 by guest on September 15, 2014http://stroke.ahajournals.org/Downloaded from

findings in a community-dwelling cohort of asymptomaticelderly individuals, and a logistic regression analysis wascarried out to define the contribution of MBs to the correctidentification of pICH patients.

Subjects and MethodsThe study population consisted of 109 patients with a final diagnosisof pICH from consecutive referrals to MRI by the neurologydepartment of a university hospital. This included all patientsadmitted with intracerebral bleeding who were able to undergo suchan examination. pICH was defined as spontaneous nontraumaticintracerebral bleeding without evidence of any kind of vascularmalformation or a brain tumor as the source of the hematoma. Thepatients ages ranged from 22 to 91 years (mean 64.9 years); therewere 59 women and 50 men. Other demographic and clinicalvariables are shown in Table 1. Hypertension was diagnosed in thecase of a past history of increased blood pressure or if the bloodpressure recordings continued to repeatedly exceed 160/95 mm Hgpast the second week after the ICH. Diabetes mellitus was defined byfasting blood sugar .140 mg% or previous treatment. Twenty-fourpatients had suffered from a previous stroke, which was reportedlyhemorrhagic in 9 patients.

MRI was performed on a 1.5-T superconducting magnet. Theimaging protocol consisted of conventional spin-echo mixed inten-sity and T2-weighted (repetition time [TR]/echo time [TE] 2300 to2600/20 to 90 ms) or fast spin-echo T2-weighted (TR 29 900 ms, TE120 ms) and cerebrospinal fluidsuppressed T2-weighted (fluid-attenuated inversion recovery; TE 130 ms, TR 6000 ms, andinversion time 1900 ms) scans, a T1-weighted (TR/TE 600/15 ms)sequence, and a gradient-echo T2*-weighted (TR/TE 600 to 800/16to 20 ms, flip angle 20 to 25) imaging series. Slice thickness wasuniformly 5 mm, and the interslice gap was 10%. All scans werereviewed by an experienced investigator who recorded the size andlocation of the clinically symptomatic hematoma and the presence ofadditional parenchymal abnormalities unaware of the patients clin-ical data. Symptomatic hematomas were grouped as (1) lobar,involving the cortex and/or deep white matter regions, (2) basalganglionic/thalamic, or (3) infratentorial, involving the brain stemand/or the cerebellum. In accordance with previous studies, MBs

were defined on gradient-echo T2*-weighted images as homoge-neous rounded lesions with a diameter of 2 to 5 mm, and theirlocation and number in specific regions of the brain was recorded(Figures 1 and 2).2,4 Larger areas of signal loss were considered torepresent old hematomas. Hyperintensities of the deep and subcor-tical white matter were specified and graded into absent, punctate,early confluent, and confluent. Periventricular hyperintensities weredescribed as caps or lining, bands, or irregular extending into the

TABLE 1. Demographic, Clinical, and Morphological Variablesfor Patients With pICH and Control Subjects

Variable pICH (N5109) Controls (N5280) P

Age, y (mean6SD) 64.9612.3 59.966.1 ,0.001*

Males, n (%) 50 (45.8) 149 (53.2) 0.19

Hypertension, n (%) 67 (61.5) 89 (31.8) ,0.001

Diabetes, n (%) 15 (13.8) 13 (4.6) 0.002

White matter hyperintensity,n (%)

Punctate 33 (30.3) 149 (53.2) ,0.001

Early confluent 27 (24.8) 28 (10)

Confluent 31 (28.4) 11 (3.9)

Periventricularhyperintensity, n (%)

Caps/lining 54 (49.5) 109 (38.9) ,0.001

Bands 15 (13.8) 11 (3.9)

Irregular 27 (24.8) 0

Lacunae, n (%) 57 (52.3) 23 (8.2) ,0.001

Infarcts, n (%) 11 (10.1) 6 (2.1) 0.002

MBs, n (%) 59 (54.1) 18 (6.4) ,0.001

Pearson x2 test was used for frequency comparisons.*Student t test.

Figure 1. Patient (aged 78 years) with pICH in left thalamusextending into basal ganglia. a and b, Fluid-attenuated inversionrecovery (TE 130 ms, TR 6000 ms, and inversion time 1900 ms)images show site of bleeding, irregular periventricular hyperin-tensity, and cribriform state of basal ganglia. c and d, Gradient-echo T2*-weighted images (TR/TE 600/16 ms, flip angle 20)reveal multiple foci of signal loss suggestive of old MBs withinthe basal ganglia and less numerous foci in cortical-subcorticalregions (arrows).

Figure 2. Patient (aged 63 years) with subacute bleeding inright parietal lobe. a, T1-weighted (TR/TE 600/15 ms) scan isshown. b and c, Gradient-echo T2*-weighted images (TR/TE600/16 ms, flip angle 20) show multiple cortical-subcortical fociof signal loss around the hematoma (arrows). d, Two MBs arealso seen in basal ganglia (arrows).

2666 Stroke November 2000

by guest on September 15, 2014http://stroke.ahajournals.org/Downloaded from

deep white matter.10 Areas of ischemic parenchymal destruction, ie,lesions exhibiting signal isointensity with cerebrospinal fluid in theircenters, were diagnosed as lacunes (,10 mm in diameter) or infarcts.

A cohort of 280 asymptomatic elderly participants of the AustrianStroke Prevention Study (ASPS) who underwent the same imagingprotocol while our study group of pICH patients was being selectedserved as a comparison. The rationale and design of the ASPS andthe MRI findings of this cohort have been published previously.7,11In short, MRI examinations with a gradient-echo T2*-weightedsequence revealed that MBs may be found even in neurologicallynormal elderly individuals. The clinical and imaging data of theASPS cohort that are relevant to the present study are summarized inTable 1.

Statistical analysis was performed with the Statistical Package ofSocial Sciences (SPSS, Inc). We used the Pearson x2 test to comparethe frequency distribution of categorical variables between groups.Continuous variables were compared with the Student t test. Alogistic regression analysis was used to define those demographic,clinical, and morphological variables that independently contributedto a correct classification of individuals as pICH patients or controls.

ResultsThe clinically symptomatic hematoma was lobar in 43 (39%)patients, it was located in the basal ganglia/thalami in 50(46%) patients, and it involved the brain stem/cerebellum in14 (13%) patients. Two patients had .1 acute bleeding atdifferent sites. MBs were seen in 59 (54%) of pICH patients,and their number ranged from 1 to 90 lesions (mean 14,median 6). Patients with MBs were significantly more oftenhypertensive and more frequently had a previous strokehistory (Table 2). Patients with MBs had significantly highergrades of deep white matter and periventricular hyperinten-sities, and they more commonly showed lacunes and largerareas of hemosiderin deposition from old hematomas (Table 3).

The majority of patients with MBs exhibited multiplelesions, which were noted simultaneously in various parts ofthe brain (Figures 1 and 2). MBs were observed in cortical-subcortical regions in 43 patients, in the basal ganglia andthalami in 41 patients, in the brain stem in 24 patients, in thecerebellum in 23 patients, and in the deep white matter in 12patients. Figure 3 shows the regional distribution of MBsaccording to the site of the acute hematoma. As can be seen,there was some tendency toward a higher frequency of MBsin the basal ganglia and infratentorial region in patients withbasal ganglionic/thalamic bleeds compared with those pa-tients with a lobar hematoma. However, these differences inthe distribution of MBs reached statistical significance onlyfor cerebellar MBs, which were seen in 1 of 43 patients witha lobar bleed and in 17 of 33 patients with a symptomatichematoma in the basal ganglia or thalami (P,0.001). Inparallel, patients with basal ganglionic/thalamic bleeds

showed a significantly greater mean number of MBs in thebasal ganglia (3.165.9 versus 0.9561.9, P,0.02) and in thecerebellum (1.062.1 versus 0.0260.15, P,0.02) than didindividuals with a lobar pICH (Figures 1 and 2). No signifi-cant differences between pICH subgroups were noted inregard to the presence and number of MBs in cortical-subcortical regions. We also did not find significant differ-ences in the distribution of MBs related to the presence orabsence of hypertension. Among the 15 patients with MBsand normal blood pressure, 12 (80%) showed MBs incortical-subcortical regions, and 8 (53%) had MBs in thebasal ganglia/thalami. MBs were seen in a cortical-subcortical location in 31 (70%) of 44 patients with hyper-tension and in the basal ganglia/thalami in 33 (75%) hyper-tensive patients with MBs.

Comparison of the pICH study group with the cohort ofneurologically asymptomatic elderly individuals showedhighly significant differences in regard to almost all clinicaland morphological variables assessed (Table 1). Therefore,we used logistic regression analysis to define those variablesthat would best allow us to correctly separate pICH patientsand control subjects. Three variables emerged that indepen-dently contributed to such classification. These variableswere the presence of MBs, the grade of periventricularhyperintensity, and the number of lacunes (Table 4). Noclinical or demographic variable entered this model.

DiscussionWe found evidence of old MBs in 54% of patients with pICHby using gradient-echo T2*-weighted MRI. This frequencycorresponds closely to that of 57% recently reported byTanaka et al5 in a smaller series of 30 patients. Earlierinvestigations that have reported a smaller rate of MBs inassociation with ICH have used conventional MRI sequencesin all or most of their patients.2,12 Therefore, these lowernumbers most likely resulted from the inferior sensitivity ofconventional T2-weighted scans in the detection of smallhemosiderin deposits. A much higher detection rate of MBshas recently been documented with the use of a gradient-echotechnique compared with conventional T2 sequences in adirect comparison.5 Greenberg et al3 observed MBs in 80% of

TABLE 2. Demographic and Clinical Variables for pICHPatients With and Without MBs

VariableWith MBs(N559)

Without MBs(N550) P

Age, y (mean6SD) 66612 63613 0.13

Males, n (%) 32 (54.2) 27 (54) 0.6

Hypertension, n (%) 44 (74.6) 23 (46) 0.003

Diabetes, n (%) 8 (13.6) 7 (14) 0.6

Previous stroke, n (%) 18 (30.5) 8 (16) 0.04

TABLE 3. Concomitant Morphological Findings for pICHPatients in Relation to Presence of MBs

VariableWith MBs(N559)

Without MBs(N550) P

White matter hyperintensity, n (%)

Punctate 14 (23.7) 19 (38) ,0.001

Early confluent 14 (23.7) 13 (26)

Confluent 28 (47.5) 3 (6)

Periventricular hyperintensity, n (%)

Caps/lining 26 (44.1) 28 (56) ,0.001

Bands 6 (10.2) 9 (18)

Irregular 25 (42.4) 2 (4)

Lacunae, n (%) 39 (66.1) 18 (36) 0.008

Infarcts, n (%) 9 (15.3) 2 (4) 0.2

Old hematoma, n (%) 24 (40.7) 5 (10) 0.001

Roob et al MRI Evidence of Microbleeds and ICH 2667


their patients, but this sample was small and was composedonly of patients with lobar hemorrhage.

The overall number of MBs in individual patients in ourseries was quite variable and ranged from 1 to 90 lesions.Most frequently, MBs were seen in cortical-subcortical re-gions or in the basal ganglia including the thalami. MBs in thebrain stem and cerebellum were less frequent, and the whitematter was relatively spared. Typically multiple MBs werefound scattered throughout the brain. Therefore, a separationof patient subgroups based on a specific location of MBs wasimpossible, and a grading of regional preferences from theabsolute number of MBs appeared problematic because ofdifferences in the size of the regions to compare (eg,cortical-subcortical versus brain stem). Therefore, we decidedto compare the distribution of MBs between patients on thebasis of the site of the symptomatic hematoma. This analysisrevealed some correspondence between pICH location andMB topography, as illustrated in Figure 3; ie, frequency andnumber of MBs in the basal ganglia/thalami and the infrat-entorial region were greater in patients with a pICH at thesesites than in patients with a lobar hematoma. However,concerning cortical-subcortical MBs, no clear differencesemerged between pICH subgroups. Thus, we were unable toconfirm a specific pattern of MBs strongly suggestive ofcerebral amyloid angiopathy in this unselected series.3 In thiscontext, the probably rather small number of patients withcerebral amyloid angiopathy in our sample has to be consid-ered. Moreover, histopathologic findings suggest a combina-tion of cerebral amyloid angiopathy with hypertensive mi-croangiopathy as the cause of a more widespread distributionof MBs in some patients.4 Interestingly, however, a prefer-

ential cortical-subcortical location of MBs was not evenfound in the normotensive pICH patients of our studypopulation.

In line with previous studies,2,5 we found highly significantassociations between the presence of MBs and other morpho-logical signs of cerebral microangiopathy, such as lacunesand extensive periventricular and deep white matter dam-age,13 and we confirmed hypertension as the single mostimportant clinical risk factor for the occurrence of MBs. Asignificantly higher frequency of a preceding stroke in pa-tients with MBs can be viewed as further clinical evidence ofa globally more pronounced vasculopathy of these patients.Comparison with a cohort of normal elderly control subjectsshowed the expected differences regarding a significantlyhigher rate of hypertension and also of diabetes mellitus inpICH patients. However, only MRI findings emerged asindependent discriminating variables in a logistic regressionmodel. These morphological variables were the presence ofMBs and the grade of periventricular hyperintensity and thenumber of lacunes. CT studies have already associatedlacunes and leukoaraiosis with a higher risk of intracerebralbleeding either spontaneously14 or after anticoagulant thera-py.15 This analysis supports MBs as a further importantmarker of bleeding-prone small-vessel diseases. In this con-text, Greenberg et al16 have recently shown the accumulationof MBs over time on repeated MRI examination of a smallgroup of patients with intracerebral bleeding. Large-scaleprospective studies using gradient-echo T2*-weighted MRIin elderly individuals are now needed to substantiate theseassumptions.

Figure 3. Frequency of MBs in different regionsof the brain in relation to site of symptomatichematoma (2 patients with .1 acute bleeding notincluded).

TABLE 4. Logistic Regression Analysis: Significant and IndependentDiscriminators Between pICH Patients and Control Subjects

Variable Regression Coefficient B Standard Error Wald P Exp (B)

MBs 1.634 0.371 19.384 0.000 5.124

PVH grade 1.275 0.231 30.389 0.000 3.580

No. of lacunes 0.958 0.314 9.322 0.002 2.608

PVH indicates periventricular hyperintensity grade.

2668 Stroke November 2000


References1. Atlas SW, Mark AS, Grossmann RI, Gomori JM. Intracranial hemor-

rhage: gradient echo MR imaging at 1.5 T: comparison with spin-echoimaging and clinical implications. Radiology. 1988;168:803807.

2. Offenbacher H, Fazekas F, Schmidt R, Koch M, Fazekas G, Kapeller P.MR of cerebral abnormalities concomitant with primary intracerebralhematomas. AJNR Am J Neuroradiol. 1996;17:573578.

3. Greenberg S, Finkelstein S, Schaefer P. Petechial hemorrhages accom-panying lobar hemorrhage: detection by gradient echo MRI. Neurology.1996;46:17511754.

4. Fazekas F, Kleinert R, Roob G, Kleinert G, Kapeller P, Schmidt R,Hartung HP. Histopathologic analysis of foci of signal loss ingradient-echo T2*-weighted MR images in patients with spontaneousintracerebral hemorrhage: evidence of microangiopathy-relatedmicrobleeds. AJNR Am J Neuroradiol. 1999;20:637642.

5. Tanaka A, Ueno Y, Nakayama Y, Takano K, Takebayashi S. Smallchronic hemorrhages and ischemic lesions in association with spon-taneous intracerebral hematomas. Stroke. 1999;30:16371642.

6. Kwa V, Franke CL, Verbeeten B Jr, Stam J. Silent intracerebral micro-hemorrhages in patients with ischemic stroke. Ann Neurol. 1998;44:372377.

7. Roob G, Schmidt R, Kapeller P, Lechner A, Hartung HP, Fazekas F. MRIevidence of past cerebral microbleeds in a healthy elderly population: theAustrian Stroke Prevention Study. Neurology. 1999;52:991994.

8. Roob G, Fazekas F. Magnetic resonance imaging of cerebral microbleeds.Curr Opin Neurol. 2000;13:6973.

9. Greenberg S. Cerebral amyloid angiopathy: prospects for clinicaldiagnosis and treatment. Neurology. 1998;51:690694.

10. Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MRI signalabnormalities at 1.5 T in Alzheimers dementia and normal aging. AJNRAm J Neuroradiol. 1987;8:421426.

11. Schmidt R, Lechner H, Fazekas F, Niederkorn K, Reinhart B, GrieshoferP, Horner S, Offenbacher H, Koch M, Eber B, Schumacher M, KapellerP, Freidl W, Dusek T. Assessment of cerebrovascular risk profiles inhealthy persons: definition of research goals and the Austrian StrokePrevention Study. Neuroepidemiology. 1994;13:308313.

12. Scharf J, Brauherr E, Forsting M, Sartor K. Significance of haemorrhagiclacunes on MRI in patients with hyperintense cerebrovascular diseasesand intracerebral haemorrhage. Neuroradiology. 1994;37:504508.

13. Fazekas F, Kleinert R, Offenbacher H, Schmidt R, Kleinert G, Payer F,Radner H, Lechner H. Pathologic correlates of incidental MRI whitematter signal hyperintensities. Neurology. 1993;43:16831689.

14. Inzitari D, Giordano GP, Ancona AL, Pracucci G, Mascalchi M,Amaducci L. Leukoaraiosis, intracerebral hemorrhage, and arterial hyper-tension. Stroke. 1990;21:14191423.

15. The Stroke Prevention in Reversible Ischemia Trial (SPIRIT) StudyGroup. A randomised trial of anticoagulants versus aspirin after cerebralischemia of presumed arterial origin. Ann Neurol. 1997;42:857865.

16. Greenberg SM, ODonnell HC, Schaefer PW, Kraft E. MRI detection ofnew hemorrhages: potential marker of progression in cerebral amyloidangiopathy. Neurology. 1999;22:11351138.

Roob et al MRI Evidence of Microbleeds and ICH 2669


frequency and location of microbleeds - roob

Documents