A

OcMSRTswpwiwClma©

K

1

cnii2T

0d

Clinical Neurology and Neurosurgery 108 (2006) 744–749

Segmental left ventricular wall motion abnormalities areassociated with lacunar ischemic stroke

Dirk Deleu a,∗, Saadat Kamran a, Ayman A. Hamad a,Samir M.K. Hamdy b, Naveed Akhtar a

a Department of Neurology, Hamad Medical Corporation, P.O. Box 3050, Doha, State of Qatarb Department of Cardiology, Hamad Medical Corporation, P.O. Box 3050, Doha, State of Qatar

Received 31 October 2005; received in revised form 10 March 2006; accepted 10 March 2006

bstract

bjective: The aim of this study was to determine whether segmental left ventricular wall motion abnormalities (SLVWMA) are a potentialause of ischemic stroke.ethods: Demographics, cardiovascular risk factors and echocardiographic parameters of patients with ischemic stroke (with and without

LVWMA) were collected and compared with those of patients who had SLVWMA but without history of ischemic stroke.esults: Two hundred and fifty nine patients with ischemic stroke were identified: 187 patients without SLVWMA, and 72 with SLVWMA.he cardiac group consisted of 79 patients. Compared with the stroke patients with SLVWMA, stroke patients without SLVWMA werelightly but significantly younger (59 versus 63 years of age). Furthermore, the number of risk factors in stroke patients without SLVWMAas significantly lower compared with stroke patients with SLVWMA (2.7 versus 3.7). There was no difference in age or gender between strokeatients with SLVWMA and the cardiac patients. However, the number of risk factors was significantly higher in the cardiac patients comparedith stroke patients with SLVWMA (4.4 versus 3.7). The ejection fraction was normal in both groups of stroke patients but significantly lower

n the cardiac patients (37%). Significantly more lacunar ischemic strokes were observed in stroke patients with SLVWMA than in thoseithout SLVWMA (76% versus 68%).onclusion: Our data indicate that in ischemic stroke patients with multiple cardiovascular risk factors and SLVWMA a higher frequency of

acunar strokes can be found. The latter could be a marker of small-vessel disease and/or be a potential contributing factor, perhaps through aechanism of cardiac microembolism, in the development of lacunar ischemic stroke. The mechanisms of the association between SLVWMA

nd lacunar ischemic stroke remain however unclear.2006 Elsevier B.V. All rights reserved.

eywords: Ischemic stroke; Lacunar stroke; Risk factor; Left ventricle; Echocardiography

taamf

. Introduction

Stroke is ranked as the third cause of death in developedountries and contributes substantially to adult long-termeurological disability [1]. Data derived from our stroke reg-stry revealed that the majority of stroke subtypes observed

n Qatar are lacunar [2]. Clinical stroke registries show that0–40% of strokes are caused by cardiac embolism [3,4].his type of stroke carries a significantly worse prognosis

∗ Corresponding author. Tel.: +974 439 27 73; fax: +974 439 18 26.E-mail address: doc deleu@hotmail.com (D. Deleu).

(QsnQt

303-8467/$ – see front matter © 2006 Elsevier B.V. All rights reserved.oi:10.1016/j.clineuro.2006.03.005

han those of alternative etiology [5] and, in addition, havehigher rate of recurrence [6]. Left ventricular hypertrophy

nd abnormal left ventricular geometry by so far unexplainedechanisms have also been identified as independent risk

actors for ischemic stroke [7–9].This study was performed in Hamad Medical Corporation

HMC) the largest single governmental hospital in the State ofatar with a capacity of 1600 beds. The estimated population

ize of Qatar is 850,000. Over the past decades, favorable eco-omic developments in the Arabian Gulf countries, includingatar, have lead to dramatic changes in life style. Similarly

o Western societies, stroke, in particular ischemic stroke,

y and N

pm

cmcrstsS

2

2

(aQtiwrHmitfsiAoo

2

2

np6umiisTenthsba

eiuapwoMdoe

2

sptpaotmoicwpTtf

2

cbmwiea

tatwtusblb

D. Deleu et al. / Clinical Neurolog

oses a major health threat and contributes substantially toorbidity and mortality.The purpose of this study was to assess the potential

ontribution of segmental left ventricular wall motion abnor-alities (SLVWMA), in the absence of any other identifiable

ardiac embolic source (e.g. ventricular aneurysm, atrial fib-illation, valvular disease, patent foramen ovale), to ischemictroke. Hence risk factor profiles and ischemic stroke sub-ypes were compared between three subgroups of patients:troke patients with and without SLVWMA, and patients withLVWMA but without stroke (cardiac patients).

. Patients and methods

.1. Setting

HMC is the only hospital in Qatar, which has facilitiesincluding two MRIs) for admitting and evaluating cardiacnd stroke patients. Free national health care service forataris is the cornerstone of the health care program. In addi-

ion, free medical service is provided for expatriates and vis-tors examined at the Accident and Emergency Department,ithout requiring any referral from a health center. Patients

egistered in the stroke and echocardiography database ofMC between January and June 2001 formed the studyaterial. The data of all patients admitted with first-ever

schemic stroke were reviewed. Since a large majority ofhese patients have multiple modifiable cardiovascular riskactors [2], echocardiography is routinely performed in ourtroke patients, which enabled us to divide our patients upnto two groups: those with and those without SLVWMA.s given in more detail later, the cardiac patients consistedf patients with SLVWMA but without any history of stroker transient ischemic attack (TIA).

.2. Data collection

.2.1. Stroke patientsDemographics, cardiovascular risk factors, history and

eurological, vascular and cardiac examination from allatients with first-ever ischemic stroke admitted during that-month period were collected. Electrocardiogram, duplexltrasonography of the carotid and vertebral arteries to deter-ine the presence of plaques or stenotic lesions, neuroimag-

ng (brain CT or MRI) with determination of: subtype ofschemic stroke (small-vessel versus large-vessel stroke) andite of stroke (hemispheric or brainstem) were reviewed.ransthoracic echocardiography (TTE) data recorded were:xistence of SLVWMA (topographical left ventricular dyski-esias, akinesia or hypokinesia) and the left ventricular ejec-ion fraction (EF). Patients with intracerebral, subarachnoidal

emorrhage, previous stroke, carotid artery plaques or steno-is on duplex ultrasonography were excluded. Since it woulde incorrect to claim a diagnosis of ischemic stroke in thebsence of neuroradiologic abnormalities, TIAs were also

lit<

eurosurgery 108 (2006) 744–749 745

xcluded. In young patients and/or when no obvious cause ofschemic stroke can be identified, screening for hypercoag-lability and hematological disorders is part of the work-upt HMC. The neuroimaging standard procedure for strokeatients at HMC consists of a brain CT scan on admission,hich is followed after a minimum of 4–6 days by a sec-nd neuroimaging, consisting in over 95% of cases of a brainRI with MR angiogram. When TTE was inconclusive (e.g.

ue to lack of cooperation of the patient, unclear imagingr high suspicion of cardioembolic source), transesophagealchocardiography (TEE) was performed.

.2.2. Cardiac patientsThe cardiac group (patients with SLVWMA but without

troke or TIA) consisting of patients admitted over the sameeriod as the stroke patients, and were randomly taken fromhe echocardiography database. Included were only thoseatients with SLVWMA but without a history of stroke, TIA,theromatous plaques at the aortic arch, or plaques or stenosisn duplex ultrasonography and with no cardiac source knowno cause embolism (e.g. atrial fibrillation, cardiac arrhyth-

ias, ventricular aneurysm, mural thrombus, patent foramenvale, valvular disease or surgery, or cardiomyopathy). Sim-larly to the stroke patients, their demographics, cardiovas-ular risk factors, history and complete physical examinationere collected, and electrocardiogram, duplex ultrasonogra-hy of the carotid and vertebral arteries, and TTE (and/orEE where deemed necessary) were reviewed. In addition,

he frequency of stroke occurrence was observed over theollowing 3 years.

.3. Diagnostic definitions and criteria

The WHO definition for stroke: “rapidly developing clini-al symptoms and/or signs resulting in focal or global distur-ance in cerebral function leading to death or persisting forore than 24 h with no apparent cause other than vascular”as used [10]. Ischemic stroke was defined as a stroke result-

ng from any cause except cardiac source known to causembolism, aortic or carotid atherosclerotic disease, hyperco-gulopathy or hematological disorder.

The following vascular risk factors were identified: hyper-ension, diabetes, dyslipidemia, smoking, obesity, coronaryrtery disease (CAD) and left ventricular systolic dysfunc-ion (LVSD). A patient was considered to be hypertensiveith an established history of blood pressure (BP) higher

han 140/90 mmHg in the non-acute phase or supervisedse of antihypertensive medication [11]. Patients with tran-ient increase in BP on admission were not considered toe hypertensive. The criteria for diagnosing diabetes mel-itus were the following: past history of supervised dia-etes control or consistently high fasting plasma glucose

evels (>7.0 mmol/L) [12]. Dyslipidemia was defined as fast-ng plasma total cholesterol levels >5.2 mmol/L, plasmariglyceride levels >2.0 mmol/L, plasma HDL-cholesterol0.9 mmol/L or plasma LDL-cholesterol >3.4 mmol/L or

7 y and N

u[ncrsynfs

(cClwasln

ictds

vbd

2

SgmecCpctts

3

TD

V

AGBN

V

E

S

BLb

46 D. Deleu et al. / Clinical Neurolog

sing supervised treatment of hypolipidemic medication.13] A body mass index (BMI) ≥ 30 kg/m2 was taken as diag-ostic criterion for obesity. Smoking was classified into twoategories: (1) non-smokers/former smokers (never smokedegularly or stopped regular smoking ≥5 years ago) and (2)mokers (regular daily cigarette smoking within the last 5ears). Because of inaccuracy of the medical records, we didot quantify the amount of consumption or attempted to dif-erentiate between cigarette, cigar and shisha (water pipe)moking. LVSD was defined as an EF < 0.45 [14].

A cerebral infarction was diagnosed when neuroimagingbrain CT scan or MRI) revealed a lesion corresponding thelinical picture. Lacunar infarcts were diagnosed when brainT scan or MRI revealed a small deep infarct (≤2 cm in the

ongest diameter) in the territory of a perforating vessel andhen the associated clinical presentation was suggestive oflacunar syndrome [15,16]. Striatocapsular infarcts, being

ub-cortical and >2 cm in diameter, were classified as non-acunar. Brain CT scans and MRIs were reviewed by botheurologist and neuroradiologist, independently.

As mentioned before, the existence of SLVWMA weredentified by echocardiography, and evaluated by at least two

ardiologists. The degree of SLVWMA was graded usinghe following criteria: akinesia was defined as less than 10%ifference in left ventricular wall size between diastole andystole, while hypokinesia referred to a difference in left

3

7

able 1emographic, clinical and neuroradiologic characteristics of ischemic stroke patien

ariable Frequency (%) or mean ± S.D. (range)

Patients with stroke withoutSLVWMA (n = 187)

PS

ge (years) 59.2 ± 12.0¶ (24–90) 6ender ratio (male/female) 2.44* 3MI (kg/m2) 31.8 ± 12.4 3umber of risk factors 2.7 ± 1.1* (median 3) 3

ascular risk factorHypertension 142 (75.9) 4Diabetes 106 (56.7) 4Dyslipidemia 111 (59.4) 3Obesity 60 (32.1) 2Smoking 44 (23.5)* 4CAD 17 (9.1)* 3LVSD 3 (1.6)* 2

jection fraction (%) 62.6 ± 6.5* 4

troke subtypeHemisph. lac. 127 (67.9)* 5Hemisph. non-lac. 16 (8.6) 5Brainst. lac. 28 (15.0) 8Brainst. non-lac. 6 (3.2) 2Comb. lac. 0 (0.0) 1Comb. non-lac./lac. 10 (5.3) 1

rainst.: brainstem; BMI: body mass index; CAD: coronary artery disease; Comb.:VSD: left ventricular systolic dysfunction; SLVWMA: segmental left ventricular waetween patients with stroke and SLVWMA.¶ p = 0.014.* p = 0.0001.

** p = 0.001.

eurosurgery 108 (2006) 744–749

entricular wall size between diastole and systole rangingetween 10–30%. Furthermore, the left ventricular wall wasivided into 16 anatomical segments.

.4. Data analysis

All data were coded using Statistical Package for Socialciences (SPSS) for Windows version 10.0 data entry pro-ram. All data for continuous variables are expressed asean ± standard deviation (S.D.) and as frequency for cat-

gorical variables. The Student’s t-test was performed toompare differences in age and EF between patient groups.omparison of non-continuous variables between groups waserformed by non-parametric testing. Pearson’s correlationoefficient was used to determine the association betweenhe left ventricular EF and the number of vascular risk fac-ors. A p-value of less than 0.05 was considered to indicatetatistically significant differences.

. Results

.1. Patient population

We identified 259 patients with stroke; 187 without and2 with SLVWMA. Seventy-nine patients fulfilled the entry

ts and cardiac patients

atients with stroke andLVWMA (n = 72)

Patients with SLVWMA withoutstroke (cardiac patients) (n = 79)

3.2 ± 10.8 (39–94) 60.9 ± 11.3 (35–81).23 3.320.1 ± 9.7 28.1 ± 3.9.7 ± 1.5 (median 4) 4.4 ± 1.3** (median 4)

9 (68.1) 62 (75.6)2 (58.3) 43 (52.4)7 (51.4) 50 (61.0)4 (33.3) 25 (30.5)7 (65.3) 40 (48.8)5 (48.6) 77 (93.9)*

7 (37.5) 61 (74.4)*

9.7 ± 12.5 37.2 ± 12.2*

5 (76.4) –(6.9) –(11.1) –(2.7) –(1.4) –(1.4) –

combination of multiple infarctions; Hemisph.: hemispheric; Lac.: lacunar;ll motion abnormalities; S.D.: standard deviation. Significance of difference

y and N

ctTco(w(

3

sop(irtbotoisS

3

coSv

3

rw(ssoS(ctsh

3

l1

4

tsdbtfiiaSeoop

wfoTfiotm[

snlditoTatmwdsi[

caetelpi

D. Deleu et al. / Clinical Neurolog

riteria in the cardiac group. The demographic data of thesehree subpopulations of patients is summarized in Table 1.he patients without SLVWMA were significantly youngerompared with those with SLVWMA (either with or with-ut stroke (cardiac patients)). Similarly, the gender ratioM/F) was significantly lower in the group of stroke patientsithout SLVWMA compared with those with SLVWMA

2.4 versus 3.2).

.2. Risk factor profile

The number of modifiable cardiovascular risk factors wasignificantly lower in the subgroup of stroke patients with-ut SLVWMA (2.7) and significantly higher in the cardiacatients (4.4) compared with stroke patients with SLVWMA3.7). The pattern of the cardiovascular risk factor profiles provided in Table 1. With regard to the type of vascularisk factors; there was no statistical difference for risk fac-ors such as hypertension, diabetes, dyslipidemia and obesityetween the three groups of patients. The median numberf vascular risk factors in the three groups was at leasthree. The type of dyslipidemia found in the three subgroupsf patients was predominantly hypercholesterolemia. Smok-ng, CAD and LVSD were significantly less frequent in thetroke patients without SLVWMA compared to patients withLVWMA (stroke patients and cardiac patients).

.3. Echocardiographic characteristics

The cardiac patients had a significantly lower EF (37%)ompared with stroke patients with SLVWMA, who hadn average a normal EF (50%) (Table 1). The presence ofLVWMA highly correlated (p = 0.001) with the number ofascular risk factors.

.4. Ischemic stroke characteristics

With regard to the location of the cerebral infarct, the ante-ior circulation (anterior and medial cerebral artery territory)as most commonly affected. Considering all stroke patients

with and without SLVWMA), the frequency of lacunar hemi-pheric infarctions was significantly higher (p = 0.0001) introke patients with low EF (<45%). Furthermore, this typef infarcts was significantly higher in stroke patients withLVWMA (76%) compared with those without SLVWMA68%) (Table 1). Eight stroke patients with SLVWMA had alinical presentation not matching the neuroradiologic loca-ion of the lacunar infarction. However, in all of them one oreveral lacunar infarcts where found elsewhere in the cerebralemispheres.

.5. Follow-up data

No stroke occurred in the cardiac patients over the fol-owing 3 years and the survival rate in the cardiac group was00%.

osti

eurosurgery 108 (2006) 744–749 747

. Discussion

To our knowledge, this is the first study, which exploreshe possible association between SLVWMA and ischemictroke. From this study, the following inferences can berawn: firstly, and as expected, an association was detectedetween CAD and SLVWMA; secondly, smoking is likelyo be associated with the development of SLVWMA; andnally, as our preliminary data indicated [17], ischemic stroke

n patients with SLVWMA was more frequently associ-ted with lacunar infarction. The latter could indicate thatLVWMA is either a possible marker of small-vessel dis-ase and/or that SLVWMA perhaps through a mechanismf cardiac microembolism contributes to the developmentf (lacunar) ischemic stroke in patients with this risk factorrofile.

Hypertension, diabetes mellitus and dyslipidemia areell-known cardiovascular risk factors for stroke and are

ound in up to 80% of patients with ischemic stroke [18]. Inur cohort, hypertension was the most prominent risk factor.here was no difference in cardiovascular risk factor pro-le, which could explain the differences in stroke subtypesbserved between the two groups of stroke patients. Consis-ent with other reports, cigarette smoking in a dose-related

anner is a known risk factor for CAD and ischemic stroke19,20].

Traditionally, small vessel occlusive disease (lacunartrokes) has been attributed to microatheroma or lipohyali-osis in the small penetrating arterioles or due to an atheromaining the parent vessel [21], and hypertension [22–24] andiabetes [23–25] have been implicated in its pathophysiolog-cal process. In contrast, dyslipidemia has no predilection forhe various subtypes of ischemic stroke [26]. The contributionf cardiac embolism to lacunar stroke has been challenged.he likelihood of an embolus taking a sharp angle to enterperforating vessel that receives such a small portion of

he cerebral blood flow has been considered unlikely. Henceost commonly patients with lacunar strokes are not pursuedith echocardiography to rule out potential sources of car-iac embolism. However the possibility that lacunar ischemictroke may occur from a cardio- or thromboembolic sources not that exceptional and should therefore not be dismissed27–29].

A recent diffusion weighted MRI—echocardiographicorrelated study provided evidence that lacunar infarctionsctually occurred in five out of nine patients with cardiacmbolism [30]. These findings together with our data seemo fuel the evidence for a cardioemboligenic hypothesis. Hith-rto, the role of cardiogenic embolism in the pathogenesis ofacunar ischemic stroke has probably been underestimated,redominantly due to the lack of concomitant TTE/TEEnvestigation in most studies. Although the mechanisms

f association between SLVWMA and lacunar ischemictroke remain unclear, echocardiography seems to be jus-ified in all stroke patients including those with lacunarnfarction. Furthermore, the findings from this study may

7 y and N

hic

ntwpstfi

ptqdmcmadlcgfsttd

sscFeFro

ssdcaan

R

[

[

[

[

[

[

[

[

[

[

[

[

[

48 D. Deleu et al. / Clinical Neurolog

ave therapeutic implications in that, patients presenting withschemic stroke and SLVWMA, could be considered for anti-oagulation.

Our study reveals that lacunar strokes were predomi-antly found in the anterior circulation, with very few ofhese events occurring in the brainstem. Similar findingsere reported by Boiten and Lodder [15]. In addition, strokeatients with SLVWMA had less brainstem lacunar ischemictrokes. However, regarding SLVWMA and stroke localiza-ion the number of patients is probably too small to draw anyrm conclusions.

The cardiovascular risk factor profile of the cardiacatients (non-stroke patients) was comparable with that ofhe stroke patients with SLVWMA, except for a higher fre-uency of CAD and lower EF. Despite this, these patientsid not develop any stroke during the following 3 years. Thisight be explained by the difference in identification pro-

ess between the stroke and non-stroke (cardiac) group. Theajority of cardiac patients (94%) had known CAD (past

cute myocardial infarction or severe angina), and was onceiagnosed, treated with antiplatelet therapy or anticoagu-ants (in case of very low EF) and their risk factors wereorrected. In contrast, SLVWMA in stroke patients wereenerally detected only during the stroke work-up. There-ore the fact that the cardiac patients did not develop anytrokes is most likely attributed to the therapeutic interven-ion. However since the cardiac patients were not subjectedo neuroimaging it cannot be ruled out whether these patientseveloped silent ischemic stroke.

Admittedly, the retrospective design of this study and themall population size of the subgroup of patients imposeome limitations, our data seem to indicate a possible asso-iation between SLVWMA and (lacunar) ischemic stroke.urthermore, the non-systematic use of TEE may have under-stimated potential sources of cardiac (micro)embolism.inally, the cardiac patients were not subjected to neu-oimaging, consequently silent strokes could not be ruledut.

In conclusion, our data indicate that patients with ischemictroke and SLVWMA have higher frequency of lacunartroke. Besides the possibility of a marker of small-vesselisease, there is a possibility that cardiac microembolismould contribute to this phenomenon. The potential mech-nisms for the latter remain however unclear. Further studiesre needed to elucidate the complex association between lacu-ar ischemic stroke and SLVWMA.

eferences

[1] Bonita R, Beaglehole R, Asplund K. The worldwide problem of stroke.

Curr Opin Neurol 1994;7:5–10.

[2] Deleu D, Hamad AA, Kamran S, El Siddig A, Al Hail H, Hamdy SMK.Ethnic variations in risk factor profile, pattern and recurrence of non-cardioembolic ischemic stroke in an Arabian Gulf state community.Arch Med Res, in press.

[

eurosurgery 108 (2006) 744–749

[3] Olsen TS, Skriver EB, Herning M. Cause of cerebral infarctionin the carotid territory: its relation to the size and the location ofthe infarct and the underlying vascular lesion. Stroke 1985;16:459–66.

[4] Bogousslavsky J, Melle GV, Regli F. The Lausanne Stroke Study:analysis of 1000 consecutive patients with first stroke. Stroke1988;19:1083–92.

[5] CE Task Force. Cardiogenic brain embolism. Arch Neurol1989;46:727–43.

[6] Comess KA, DeRook FA, Beach KW, et al. Transesophageal echocar-diography and carotid ultrasound in patients with cerebral ischemia:prevalence of findings and recurrent stroke risk. J Am Coll Cardiol1994;23:1598–603.

[7] Di Tullio MR, Zwas DR, Sacco RL, Sciacca RR, Homma S. Left ven-tricular mass and geometry and the risk of ischemic stroke. Stroke2003;34:2380–6.

[8] Shaper AG, Phillips AN, Pocock SJ, Walker M, Macfarlane PW.Risk factors for stroke in middle aged British men. BMJ 1991;302:1111–5.

[9] The Dutch TIA Trial Study Group. Trial of secondary prevention withatenolol after transient ischemic attack or nondisabling ischemic stroke.Stroke 1993;24:543–8.

10] World Health Organization. Cerebrovascular diseases: prevention,treatment, and rehabilitation: report of a WHO meeting. World HealthOrgan Tech Rep Ser No 469, 1971.

11] MacMahon S, Rodgers A. The epidemiological association betweenblood pressure and stroke: implications for primary and secondary pre-vention. Hypertens Res 1994;17(Suppl 1):S23–32.

12] Anonymous. Report of the Expert Committee on the diagnosis andclassification of diabetes mellitus. Diab Care 2000;25(Suppl 1):S5–S20.

13] Expert panel on detection, evaluation and treatment of high bloodcholesterol in adults. Executive summary of the third report ofthe national cholesterol education program (NCEP) Expert panelon detection, evaluation, and treatment of high blood choles-terol in adults (adult treatment panel III). JAMA 2001;285: 2486–97.

14] Talreja D, Gruver C, Sklenar J, Dent J, Kaul S. Efficient utilization ofechocardiography for the assessment of left ventricular systolic func-tion. Am Heart J 2000;139:394–8.

15] Boiten J, Lodder J. Lacunar infarcts. Pathogenesis and validity of theclinical syndromes. Stroke 1991;22:1374–8.

16] Wardlaw JM. What causes lacunar stroke? J Neurol Neurosurg Psychi-atry 2005;76:617–9.

17] Deleu D, Kamran S, Hamad AA. Are segmental left ventricular wallmotion abnormalities a risk factor for large vessel stroke? J Neurol2005;252(Suppl 2):76.

18] Sandercock PA, Warlow CP, Jones LN, Starkey IR. Predisposing factorsfor cerebral infarction: the Oxfordshire community stroke project. BMJ1989;298:75–80.

19] Wolf PA, D’Agostino RB, Kannel WB, Bonita R, Belanger AJ.Cigarette smoking as a risk factor for stroke. The Framingham Study.JAMA 1988;259:1025–9.

20] Cavender JB, Rogers WJ, Fisher LD, Gersh BJ, Coggin CJ, Myers WO.Effects of smoking on survival and morbidity in patients randomizedto medical or surgical therapy in the Coronary Artery Surgery Study(CASS): 10-year follow-up. CASS Investigators. J Am Coll Cardiol1992;20:287–94.

21] Fisher CM. Lacunar strokes and infarcts: a review. Neurology1982;32:871–6.

22] Arboix A, Morcillo C, Garcia-Eroles L, Oliveres M, Massons J, TargaC. Different vascular risk factor profiles in ischemic stroke subtypes:

a study from the “Sagrat Cor Hospital of Barcelona Stroke Registry”.Acta Neurol Scand 2000;102:264–70.

23] Mast H, Thompson JL, Lee SH, Mohr JP, Sacco RL. Hypertension anddiabetes mellitus as determinants of multiple lacunar infarcts. Stroke1995;26:30–3.

y and N

[

[

[

[

[

[

Stroke 2002;33:2019–24.

D. Deleu et al. / Clinical Neurolog

24] Sacco SE, Whisnant JP, Broderick JP, Philips SJ, O’Fallon WM. Epi-demiological characteristics of lacunar infarcts in a population. Stroke1991;22:1236–41.

25] Pinto A, Tuttolomondo A, Di Raimondo D, Fernandez P, Licata G.Cerebrovascular risk factors and clinical classification of strokes. SeminVasc Med 2004;4:287–303.

26] Yip PK, Jeng JS, Lee TK, Chang YC, Huang ZS, Ng SK, et al. Subtypesof ischemic stroke. A hospital-based stroke registry in Taiwan (SCAN-IV). Stroke 1997;28:2507–12.

27] Kumral E, Evyapan D, Balkir K. Acute caudate vascular lesions. Stroke1999;30:100–8.

[

eurosurgery 108 (2006) 744–749 749

28] Ay H, Oliveira-Filho J, Buonanno FS, et al. Diffusion-weighted imagingidentifies a subset of lacunar infarction associated with embolic source.Stroke 1999;30:2644–50.

29] Gerraty RP, Parsons MW, Barber A, et al. Examining the lacunarhypothesis with diffusion and perfusion magnetic resonance imaging.

30] Seifert T, Enzinger C, Storch MK, Pichler K, Niederkorn K, FazekasF. Acute small subcortical infarctions on diffusion weighted MRI:clinical presentation and aetiology. J Neurol Neurosurg Psychiatry2005;76:1520–4.