Embed Size (px)
Citation preview
J A C C : C A R D I O V A S C U L A R I M A G I N G V O L . 5 , N O . 4 , 2 0 1 2
© 2 0 1 2 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 1 9 3 6 - 8 7 8 X / $ 3 6 . 0 0
P U B L I S H E D B Y E L S E V I E R I N C . D O I : 1 0 . 1 0 1 6 / j . j c m g . 2 0 1 1 . 1 0 . 0 0 7
Plaque Features Associated With IncreasedCerebral Infarction After Minor Stroke and TIAA Prospective, Case-Control, 3-T Carotid Artery MR Imaging Study
Alistair C. Lindsay, MBCHB, DPHIL,* Luca Biasiolli, MSC,* Justin M. S. Lee, MA,*Ilias Kylintireas, MD, DPHIL,* Bradley J. MacIntosh, PHD,† Hilary Watt, MA, MSC,‡Peter Jezzard, PHD,† Matthew D. Robson, PHD,* Stefan Neubauer, MD,*Ashok Handa, MD,§ James Kennedy, MD,� Robin P. Choudhury, MA, DM*
Oxford, United Kingdom
O B J E C T I V E S The goal of this study was to determine whether a 3-T magnetic resonance imaging
(MRI) protocol combining carotid atherosclerotic plaque and brain imaging can identify features of
high-risk acutely symptomatic plaque that correlate with brain injury.
B A C K G R O U N D It has previously been demonstrated that, in asymptomatic patients, MRI can identify
features of carotid plaque that are associated with stroke, such as the presence of a large lipid core. We
hypothesized that the early phase (�7 days) after a cerebrovascular event, when risk of recurrence is highest,
may be associated with particular plaque characteristics that associate with cerebral injury.
M E T H O D S Eighty-one patients (41 presenting acutely with transient ischemic attack [TIA] or minor
stroke and 40 asymptomatic controls) underwent multicontrast carotid artery MRI on 2 separate
occasions, each accompanied by diffusion-weighted imaging (DWI) and fluid-attenuated inversion
recovery (FLAIR) imaging of the brain.
R E S U L T S Complex (American Heart Association [AHA] type VI) plaques were seen in 22 of 41
patients (54%) in the symptomatic group versus 8 of 40 (20%) in the asymptomatic group (p � 0.05).
They were caused by intraplaque hemorrhage (34% vs. 18%; p� 0.08), surface rupture (24% vs. 5%; p� 0.03),
or luminal thrombus (7% vs. 0%; p � 0.24). Noticeably, 17 of 30 (57%) cases of AHA type VI plaque were
in vessels with �70% stenosis. At follow-up scanning (�6 weeks later), only 2 cases of symptomatic AHA
type VI plaque showed evidence of full healing. The presence of fibrous cap rupture was associated with
higher DWI brain injury at presentation and higher total cerebral FLAIR signal at follow-up (p � 0.05).
C O N C L U S I O N S Early carotid wall MRI in patients experiencing minor stroke or TIA showed a higher
proportion of “complex” plaques compared with asymptomatic controls; a majority were in arteries of �70%
stenosis. Fibrous cap rupture was associated with increases in DWI and FLAIR lesions in the brain. Combined
carotid plaque and brain MRI may aid risk stratification and treatment selection in acute stroke and
TIA. (J Am Coll Cardiol Img 2012;5:388–96) © 2012 by the American College of Cardiology Foundation
From the *Department of Cardiovascular Medicine, University of Oxford, Oxford, United Kingdom; †Nuffield Department of
Clinical Neurosciences, University of Oxford, Oxford, United Kingdom; ‡Centre for Statistics in Medicine, University of
Srssccman
repetition time
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 5 , N O . 4 , 2 0 1 2
A P R I L 2 0 1 2 : 3 8 8 – 9 6
Lindsay et al.
Acute 3-T MRI of Carotid Plaque in Stroke
389
troke is a major cause of mortality anddisability (1). After an initial transient isch-emic attack (TIA) or minor stroke second-ary to large artery atherosclerosis, patients
emain at a particularly high risk of disablingtroke within the first 7 days (2). Patients withevere (70% to 99%) ipsilateral carotid stenosis onarotid ultrasound are currently treated with earlyarotid endarterectomy, and some patients withoderate (50% to 69%) symptomatic stenosis may
lso benefit from surgery (3,4). However, determi-ation of individual patient risk and selection of
See page 406
optimal therapy remain challenging. The associa-tion between carotid artery stenosis and stroke riskwas originally demonstrated using x-ray arteriogra-phy (5,6). This technique is invasive, relativelyexpensive, has limited availability, and has largelybeen replaced by noninvasive imaging with Dopplerultrasound, computed tomography (CT) angiogra-phy, and magnetic resonance (MR) angiography(7). However, because atherosclerotic plaques accu-mulate within the vessel wall and are heterogeneousin both composition and morphology, angiographictechniques cannot reveal the full complexity ofplaque disease, and opportunities for more precisediagnosis and stratification of risk, based on appre-ciation of plaque characteristics, may be overlooked(8). This limitation of conventional angiographymay be of particular importance in moderately sizedplaques with modest impingement on the vessellumen (9).
Magnetic resonance imaging (MRI) is noninva-sive, does not involve harmful ionizing radiation,provides high-resolution images, and is widely avail-able for clinical use. The ability of MRI to accuratelyidentify potential features of “high-risk” or “unstable”plaque such as hemorrhage (10), surface rupture (11),lipid-rich necrotic core (10), and thrombus (12) hasnow been extensively validated by comparison of
Oxford, Oxford, United Kingdom; §Nuffield Department of SurgicalSciences, University of Oxford, Oxford, United Kingdom; and the�Nuffield Department of Medicine, University of Oxford, Oxford, UnitedKingdom. Dr. Choudhury is a Wellcome Trust Senior Research Fellowin Clinical Science. Dr. Lindsay was a Radcliffe Cardiovascular ResearchFellow. This study was supported by the Oxford Comprehensive Bio-medical Research Centre, National Institute for Health Research fundingscheme. Drs. Neubauer and Choudhury acknowledge the support of theBritish Heart Foundation Centre of Research Excellence, Oxford. Allother authors have reported that they have no relationships relevant to thecontents of this paper to disclose.
Manuscript received October 7, 2011; accepted October 17, 2011.
imaged plaques with histological examination ofcarotid endarterectomy samples (10,12–15).
A MR classification of plaque, modified from theAmerican Heart Association (AHA) grading sys-tem has been developed. Type IV/V plaque de-scribes “vulnerable” plaque with a large lipid core,whereas AHA type VI plaque exhibits featuresconsistent with acute events, as suggested by intra-plaque hemorrhage, cap rupture, or thrombus (16).
A number of studies have evaluated carotidplaque with MRI after stroke, but recruitment tothe majority of these studies was extended up to 3 to12 months after the index event so that plaquefeatures very early after the event (relating to thetime of greatest risk and possibly the source ofinfarction) remain largely unexplored (17–21). Al-ternatively, carotid arteries may have been imagedearly after an acute event but without acontrol group or follow-up scan for com-parison (22) or in small numbers withoutquantification of the extent of brain injury(23).
Accordingly, the present study investi-gated the ability of high-resolution MRIat 3-T to characterize carotid plaques ofpatients early after minor stroke or TIA. Todetermine the natural history of symptom-atic plaque, patients were re-examined withcarotid MRI after 6 weeks. MRI-derivedplaque characteristics were compared with astable matched population of asymptomaticpatients. Lastly, to test whether “acute”carotid plaque characteristics related tocerebral infarction, diffusion-weighted im-aging (DWI) (for acute lesions) (24) andfluid-attenuated inversion recovery le-sions (FLAIR) (for established infarcts)(25) in the brain on MRI were quantified in bothpopulations.
M E T H O D S
Patient recruitment. This was a single-center studyin a tertiary hospital. The protocol was approved bythe local research ethics committee, and all patientsprovided written informed consent.
Potential study participants were identified frompatients presenting with symptoms of a first TIA orstroke to the Acute Stroke Service at the JohnRadcliffe Hospital, Oxford, United Kingdom. Allparticipants had the diagnosis of a cerebrovascularevent confirmed by 2 independent physicians. Con-
A B B
A N D
CT �
DWI �
imagi
FLAIR
invers
MR �
MRI �
imagi
PDW
densi
T-1W
T-2W
TE �
TIA �
TOF �
TR �
secutive patients with an ipsilateral carotid
R E V I A T I O N S
A C R O N YM S
computed tomography
diffusion-weighted
ng
� fluid-attenuated
ion recovery
magnetic resonance
magnetic resonance
ng
� proton
ty–weighted
� T-1–weighted
� T-2–weighted
echo delay time
transient ischemic attack
time of flight
artery
SruAdcabJoaMPe
plaiacoD(tepnwsrtiiAm
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 5 , N O . 4 , 2 0 1 2
A P R I L 2 0 1 2 : 3 8 8 – 9 6
Lindsay et al.
Acute 3-T MRI of Carotid Plaque in Stroke
390
stenosis of �30% (according to North Americanymptomatic Carotid Endarterectomy Trial crite-ia) on screening ultrasound were recruited andnderwent the study MRI protocol as inpatients.symptomatic patients matched primarily for theegree of carotid stenosis (as defined by usingarotid ultrasound), and subsequently also by agend sex, were recruited as controls from the data-ase of the vascular ultrasound department at theohn Radcliffe Hospital and underwent MRI asutpatients. Patients with acute stroke or TIA werepproached as soon as clinically appropriate and
RI obtained as quickly as possible thereafter.atients who had suffered an acute cerebrovascularvent �1 week beforehand were not recruited. To
maximize the likelihood of the carotid artery beingthe cause of the stroke/TIA, patients with atrialfibrillation or a recent acute coronary syndromewere excluded. Patients were scanned at the OxfordCentre for Clinical Magnetic Resonance using aSiemens Trio 3T MRI scanner (Siemens, Erlangen,Germany).Carotid wall MRI. All patients were imaged using a
hased-array 4-channel carotid coil. After an initialocalizer sequence, a time-of-flight (TOF) MRngiography sequence obtained bright-blood imag-ng of each carotid artery. The suspected culpritrtery (i.e., the artery contralateral to the side oflinical symptoms) was selected, and the bifurcationf this vessel was used as the center landmark.ark-blood T-1–weighted (T-1W), T-2–weighted
T-2W), and proton density–weighted (PDW)urbo spin echo images were then taken to 10 mmither side of the culprit carotid bifurcation (in-lane resolution 0.47 mm � 0.47 mm; slice thick-ess 2 mm). Subsequent T-2W and PDW slicesere then obtained from any atherosclerotic plaque
een. Scan protocols were as follows: for T-1W,epetition time (TR) of 1 R-R interval, echo delayime (TE) of 14 ms; for T-2W, TR of 2 R-Rnterval, TE of 89 ms; for PDW, TR of 2 R-Rnterval, TE of 14 ms; for TOF, TR/TE 72/4.1 ms.ll turbo spin echo images were obtained with aatrix size of 320 � 320.
Brain MRI. After carotid wall imaging, MR brainimaging data were collected using a 12-channelhead receive coil. After an initial localizer sequence,all patients underwent DWI (TR/TE � 4,436/93ms, b values � 0.1000 s/mm2, 27 slices 4.5 mmapart, voxel size 1.6 � 1.6 �3 mm) and FLAIR(TR/TE/inversion time � 13,970/94/2,500 ms; FA� 150 degrees, 27 slices 4.5 mm apart; voxel size 0.9
� 0.9 � 3 mm) imaging of the brain.Image analysis. All MR images from each of thecontrast weightings were reviewed by the sameoperator (A.C.L.) and an independent reviewerexperienced in MR plaque interpretation (L.B.).The reviewers were blinded as to the identificationand clinical data for each patient. Where interpre-tations differed, the final classification was decidedon by consensus after a further combined review ofthe images (n � 6).
Plaques were assigned a grade of I to VIIIaccording to the modified AHA plaque gradingsystem for MRI (11). In all cases of surface disrup-tion, breach of integrity of the fibrous cap wasconfirmed as described by Hatsukami et al. (11).Wall measurements were obtained using semiauto-mated image quantification software developed atthe University of Oxford and OsiriX imaging soft-ware (26).
All brain imaging analysis was performed usingOsiriX and Jim 5.0 imaging analysis software(Xinapse Systems Ltd., Aldwincle, United King-dom) by a single reader who was blinded toplaque imaging results and clinical symptoms(A.C.L.). Outlining of all pathological lesionsseen on axial DWI and FLAIR imaging of thebrain was performed for each slice and convertedinto a total lesion volume for each hemisphere.To correct for intersubject variability and coil-related signal inhomogeneities, all images weremanually thresholded so that all lesions werecompared with the signal intensity of whitematter with normal appearance in each hemi-sphere and for each axial section.Statistical analysis. The Fisher exact test was used totest the baseline characteristics between theacutely symptomatic group and the control co-hort. In addition, comparison of plaque gradesbetween cohorts was conducted by using theWilcoxon signed rank test in addition to theFisher exact test. DWI data were nonparametri-cally distributed, and a Mann-Whitney U testwas used to compare cerebral injury values be-tween groups that were divided by carotid plaquecharacteristics.
R E S U L T S
Eighty-two patients were recruited. Of the 42symptomatic patients, 1 was unable to tolerate theentire scan procedure; therefore, images were avail-able from 41 symptomatic patients for analysis. Fullimaging data were obtained in all 40 patients in the
asymptomatic cohort. The groups were well
scssiaTelt
f9p
w
ocsrhua(aiw
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 5 , N O . 4 , 2 0 1 2
A P R I L 2 0 1 2 : 3 8 8 – 9 6
Lindsay et al.
Acute 3-T MRI of Carotid Plaque in Stroke
391
matched for clinical characteristics and percentnarrowing of the artery measured by ultrasound(Table 1). All patients in the symptomatic cohortreceived a final diagnosis of TIA or stroke. Afterreview by 2 independent physicians, 24 patientswere given a final diagnosis of stroke and 17 weregiven a final diagnosis of TIA according to WorldHealth Organization criteria (27). The median timefrom symptom onset to MRI in the symptomaticgroup was 2.1 days (range 0.17 to 7.0 days). Allpatients were treated with aspirin and statin ther-apy.Plaque analysis. The distribution of lesion typesbetween the culprit and control arteries is shownin Figure 1. Features consistent with intraplaquehemorrhage, fibrous cap rupture, or mural throm-bus (AHA type VI plaque) (Fig. 2) were seen in22 of 41 (54%) symptomatic patients but in only8 of 40 (20%; p � 0.01) asymptomatic patients.Significantly, 12 of 22 (55%) AHA type VIplaques in the symptomatic group were in vesselswith �70% stenosis of the lumen, as assessed byultrasound.AHA type VI plaque analysis. Surface rupture waseen at all degrees of stenosis, including 4 casesausing �70% stenosis and 2 cases causing �50%tenosis of the carotid artery, as assessed by ultra-ound (Fig. 3). Surface rupture was more commonn symptomatic carotid arteries compared withsymptomatic arteries (24% vs. 5%; p � 0.03).hree patients in the symptomatic cohort showed
vidence of mural thrombus within the arterialumen, a feature that was not seen in the asymp-omatic cohort (7% vs. 0%; p � 0.24). Two of these
cases were associated with stroke and with luminalobstruction of �70% on carotid ultrasound, and thethird case was associated with TIA and an ultra-sound stenosis of only 30%. Intraplaque hemor-rhage (characterized by bright signal on T-1, T-2,and TOF) was seen in 13 of 22 cases of AHA typeVI plaque in the acutely symptomatic group and in7 of 8 cases of AHA type VI plaque in theasymptomatic cohort.Follow-up imaging. All patients in the acute cohortwere invited for follow-up MRI, of which 30 of 41(73%) attended. The mean time to follow-up scanwas 90 days (range 30 to 250 days; SD 48 days).The mean � SD age of patients undergoing theollow-up MRI was 73.8 � 9.9 years (range 48.0 to0.6 years), and the mean Barthel index at firstresentation was 83.8.
Follow-up plaque analysis. Of the 30 patients who
ere scanned at follow-up, 14 had shown evidencef acute plaque rupture (AHA type VI) in theulprit artery at their initial scan, 6 of whom wereubsequently treated by endarterectomy. Of theemaining 8 AHA type VI plaques scanned, 6 stillad AHA type VI plaque characteristics at follow-p, although 2 plaques showed evidence of healingnd were reclassified as AHA type IV/V plaqueFig. 4). No plaque reclassification was necessary inny other follow-up scans (i.e., all other plaquesnitially classified as AHA types III, IV/V, and VIere unchanged).
MRI assessment of brain injury. Of the 41 patients inthe acute group, evidence of cerebral injury onDWI was seen in 32 of 41 patients; the mediannumber of lesions per patient was 7, and themedian total lesion volume was 10.62 ml (range 0to 522 ml). Overall, no significant associationswere noted between AHA plaque type and down-stream cerebral injury (Table 2). However, thepresence of fibrous cap rupture was associatedwith a higher total DWI burden at presentation,and a higher number of DWI lesions overall,compared with all other plaque features (p � 0.05 forboth) (Table 2, Fig. 5). In addition, patients withsurface rupture also had higher total cerebralFLAIR signal at follow-up (p � 0.05) (Table 2).No DWI lesions were noted in the control group,and no significant associations were found in thecontrol group between plaque characteristics andFLAIR signal.
D I S C U S S I O N
This study used MRI at 3-T to undertake com-bined carotid artery plaque and brain imaging inpatients presenting acutely with minor stroke or
Table 1. Clinical Characteristics of Symptomatic and Asymptom
Symptomatic(n � 40)
Asympto(n �
Female 13 11
Age (yrs) 75.8 � 11.4 72.3 � 9
Smoking history 23 28
Hypertension 30 34
Diabetes 11 8
Previous myocardial infarction 8 9
Previous CABG 4 5
Statin treatment 26 31
Previous atrial fibrillation 2 2
Mean stenosis % (ultrasound) 59 � 21.9% 57 � 2
Values are n or mean � SD.CABG � coronary artery bypass graft.
atic Patients
matic41) p Value
0.52
.2 0.15
0.29
0.20
0.34
0.75
0.70
0.05
0.98
2.4% 0.53
TIA. In this group of high-risk patients, features of
.
Abbreviati
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 5 , N O . 4 , 2 0 1 2
A P R I L 2 0 1 2 : 3 8 8 – 9 6
Lindsay et al.
Acute 3-T MRI of Carotid Plaque in Stroke
392
complex (AHA type VI) plaque were found morefrequently than in a matched cohort of patients withasymptomatic carotid artery disease. Specifically,fibrous cap rupture was more common in thesymptomatic group and was associated with more
Asymptomatic
IV/V(16)
III(8)
Hemorrhage(6)
Rupture(2)VIII
(3)
VII(5)
VI(8)
Figure 1. Distribution of Plaque Type According to Modified AHPresentation as Asymptomatic and Symptomatic
Red segments indicate American Heart Association (AHA) type VI (ccommonly due to cap rupture. MRI � magnetic resonance imaging
MRI Appearances of Plaque With Gross Anatomical andic Correlates
aque hemorrhage into a 90% stenosis of the left internalery on high-field microscopy. The gross specimen in the mid-onfirms hemorrhage, with erythrocytes in the body of themicroscopy (hematoxylin and eosin [H&E] stain). (B) Imageixed thrombus providing a speckled appearance on MRI; grossl and histological analyses revealed both old and new throm-tion and some calcification. (C) Image shows a lipid-rich plaqueear signs of rupture on MRI, which is confirmed by histologicalat showed an intact fibrous cap and cholesterol crystals.
ons as in Figure 1.extensive downstream brain injury (quantified byusing DWI). Although current treatment algo-rithms incorporate measures of the degree of ste-nosis, the majority of cases of symptomatic AHAtype VI plaque were associated with �70% luminalstenosis on ultrasound, the current cutoff for surgi-cal endarterectomy. Compared with previous stud-ies, the current investigation combines importantnew insights gained from 4 distinctive approaches:1) patients were imaged in the first hours ofpresentation (i.e., close to the index event, corre-sponding to the period of highest risk of recur-rence); 2) a “stable” control group matched forclinical indices and for degree of luminal stenosiswas included; 3) follow-up scans were conducted todefine changes over time; and 4) the pathologicalsignificance of the MR features of carotid plaquewas investigated by quantification of downstreambrain injury.
A majority of patients with acute neurologicalsymptoms showed evidence of plaque instability inthe ipsilateral carotid artery (AHA type VI plaque[54%]). A further large proportion of symptomaticpatients had plaque that demonstrated “vulnerable”or “high-risk” features but no signs of acute rupture(AHA types IV/V [29%]). These findings are inkeeping with earlier histological analyses of plaquesremoved at carotid endarterectomy. For example,Redgrave et al. (28) analyzed 526 carotid plaquesfrom patients undergoing endarterectomy afteracute neurological symptoms and found that 59% ofpatients had ruptured plaque, similar to the rate
Symptomatic
III(6)
Hemorrhage(9)
Rupture(10)
Thrombus(3)
VII(1)
VI(22)
IV/V(12)
lassification for MRI and Divided According to Clinical
lex) plaque, which predominates in the symptomatic group, most
A C
omp
Figure 2.Microscop
(A) Intraplcarotid artdle row cplaque onshows a manatomicabus formawith no clanalysis th
seen here. In patients undergoing carotid endarter-
rCtpiMmett
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 5 , N O . 4 , 2 0 1 2
A P R I L 2 0 1 2 : 3 8 8 – 9 6
Lindsay et al.
Acute 3-T MRI of Carotid Plaque in Stroke
393
ectomy, with preoperative carotid MRI, Yuan et al.(15) described an increased incidence of MRI-defined fibrous cap rupture of 70%; patients inwhom ruptured cap was identified were 23 timesmore likely to have suffered a TIA or stroke in thepreceding 90 days.
Studies of patients undergoing carotid endarter-ectomy are inevitably biased toward severely ste-nosed arteries. Although it is well established thatdegree of carotid stenosis predicts risk when thestenosis is �70%, the relationship between strokeisk and lesser degrees of stenosis is weaker (29).arotid ultrasound imaging does not provide de-
ailed information on plaque morphology or com-osition, and it is possible that important pathologyn mild to moderately stenotic plaque is overlooked.
ore refined characterization of the arterial wallay allow stratification of risk and suggest differ-
ntial treatment pathways accordingly. In symp-omatic patients in the present study, 55% of AHAype VI plaques were found in lesions of �70%
stenosis. These findings are consistent with therecent report of Parmar et al. (22), who found thattype VI plaque identified with acute MRI at 1.5-Twas associated with ipsilateral TIA and ischemicstroke.
The present study for the first time uses MRIsystematically to connect specific features ofacutely symptomatic carotid plaque with burdenof brain injury. The association of ruptured fi-brous cap and downstream infarction suggeststhat stratification of patients on the basis of MRIplaque characteristics may be of benefit in futureintervention trials, as has been suggested byprevious histological (30) and ultrasound (31)studies. Long-term studies are underway that aimto understand the association between features ofplaque vulnerability on MRI and the risk offuture cardiovascular events, independent of thedegree of luminal stenosis (18).
The natural history of acutely symptomatic ca-rotid plaque was evaluated by using interval imag-ing. Changes over time have previously been in-ferred by comparing features such as cap rupture,macrophage content, lipid-rich core, and calcifi-cation in endarterectomy samples obtained frompatients at various time intervals after an acuteevent (28). To the best of our knowledge, this hasnot been attempted previously by serial imaging.Strikingly, even with the high-resolution andsignal-to-noise ratio offered by 3-T MRI, a largemajority of vulnerable (AHA type IV/V) and
unstable (AHA type VI) plaques did not showany evidence of change in the initial weeks afteracute stroke or TIA. This finding is consistentwith the persistence of increased risk of recur-rence of same-territory stroke within the firstmonth of acute first presentation and may suggestthat a ruptured plaque remains as a nidus for
Figure 3. Multicontrast Appearances of Acute Plaque in StenoseStenosed Carotid Arteries
(A) Evidence of acute plaque rupture in patients with �70% carcontiguous slices shown). This plaque was estimated at 30% stensound. Rupture of a thin fibrous cap can be seen (left column),have led to thrombus propagating into the lumen (right column50% stenosis, which seems to have a large surface defect. Time-imaging confirmed fibrous cap rupture (due to the presence of bplaque body) and also calcification.
Figure 4. Minimal Change in Plaque Classification Between AcuScans (T-1 Images)
(A) Bottom panel demonstrates the healing of a small surface(B) Images show a large surface defect that seems to have partly renot healed. (C) Images show bright signal suggestive of intraplaque
d and Minimally
otid stenosis (2osis on ultra-which seems to). (B) Image showsof-flight (TOF)right blood in the
te and Follow-Up
ulceration (arrow).modelled but hashemorrhage,
which displays no change at follow-up.
case was associated w
�
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 5 , N O . 4 , 2 0 1 2
A P R I L 2 0 1 2 : 3 8 8 – 9 6
Lindsay et al.
Acute 3-T MRI of Carotid Plaque in Stroke
394
thrombus formation in the early stages. The dataare also consistent with a previous study in whichonly 4 of 28 patients with ruptured coronaryartery plaques demonstrated healing after 1 year,as assessed by using intravascular ultrasound (32).Although plaque rupture is thought to be onemethod by which atherosclerotic plaque diseaseprogresses (33), the current data suggest that suchstepwise progression does not occur commonly inthe short term. Further long-term follow-up willbe needed to determine if some ruptured plaquesgo on to heal at a later stage.
Although ultrasound (34) and CT (35) studieshave attempted to link carotid plaque characteristicswith the extent of cerebral injury in patients withstroke, these studies have not been conducted dur-ing the acute phase of presentation. Furthermore,
of Plaque Changes and Downstream Injury
-1 weighted images; the bottom row shows DWI images in theere. A and C are looking at the left carotid. (A) Top panel showsrrhage associated with minimal diffusion-weighted imaging (DWI)) in the left anterior lobe. (B) A large thrombus is seen in theternal carotid artery, which was associated with only minimalanterior lobe. (C) Clear surface disruption is seen, which in this
Table 2. Acute (DWI) and Chronic (FLAIR) Downstream Damage
Plaque Gr
Imaging MethodIII
(n � 6)IV/V
(n � 12
DWI, acute; median number of lesions 7.5 (3–11) 4.0 (3–9
DWI, acute; total volume (ml) 3.2 2.4
FLAIR, follow-up; total volume (ml) 5.0 2.3
Values are median (interquartile range [IQR]) or geometric means. *p � 0.05 foAHA � American Heart Association; DWI � diffusion-weighted imaging; FLAIR
ith a large infarct in the left cerebral hemispheres.
these modalities cannot provide the detailed plaquecharacterization afforded by 3-T MRI. The presentstudy related the detailed plaque findings withquantitative measures of acute and chronic cerebralinjury. No relationship could be established be-tween the presence of conventionally defined AHAtype VI plaque and the extent of ipsilateral cerebralinjury, in either the acute or chronic phase. How-ever, when acute infarcts were quantified by usingDWI, the presence of ruptured fibrous cap wasassociated with both an increased number of DWIlesions and a greater total DWI load. Surfacerupture was also associated with a higher overallvolume of established infarcts quantified by FLAIRinjury at follow-up.Study limitations. Twenty-seven percent of patientsin the acute group were unable to attend forfollow-up scans. Of these patients, the mean agewas higher (83 years; p � 0.05) and the meanBarthel score at presentation was lower (55; p �0.01) than those who did return. Thus, those whowere unable to return were more elderly and hadexperienced more severe disability after their strokeand were unable to undergo what was an electivefollow-up study.
At the commencement of this real-world study,administration of gadolinium-based contrast agentswas not permitted because the majority of patientspresented with an estimated glomerular filtrationrate �60 ml/min. A previous study by Cai et al.(36) found that use of gadolinium contrast canenhance the definition of intact fibrous cap, al-though that study deliberately excluded patientswith fibrous cap rupture. We used the parametersdescribed by Hatsukami et al. (10), which do notinclude the use of gadolinium contrast, to confirmfibrous cap rupture. Therefore, it is possible thatdetection of fibrous cap rupture may have beenfurther improved by the use of gadolinium contrast,although its use is likely to be contraindicated in asignificant number of elderly patients with estab-
ording to AHA Plaque Grade and Plaque Characteristics
Plaque Characteristics
VI(n � 22)
Hemorrhage(n � 13)
Cap Rupture(n � 10)
Thrombus(n � 3)
6.0 (1–16) 3.0 (1–7) 16.5* (7–17) 2.0 (2–12)
2.2 1.0 4.9* 1.2
2.0 0.8 4.6* 7.1
mparison with all other plaque types.fluid-attenuated inversion recovery.
Figure 5. Examples
The top row shows Tleft cerebral hemisphan intraplaque hemoinjury (bottom panellumen of the right indamage in the right
Acc
ade
)
)
r co
lished vascular disease.
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 5 , N O . 4 , 2 0 1 2
A P R I L 2 0 1 2 : 3 8 8 – 9 6
Lindsay et al.
Acute 3-T MRI of Carotid Plaque in Stroke
395
C O N C L U S I O N S
Use of high-resolution MRI at 3-T to characterizecarotid plaques of patients within 7 days of minorstroke or TIA was feasible and showed a higherproportion of complex AHA type VI plaques com-pared with asymptomatic control patients. Greaterthan 50% of complex plaques were found in arteries
cores and intraplaque hemorrhage in
1
1
1
1
1
1
1
1
1
20. Altaf N, Daniels L,Detection of intrap
associated with increases in both DWI and FLAIRlesions in the brain. These findings may provide abasis for stratification of patients acutely accordingto lesion type and not merely by the extent ofluminal narrowing.
Reprint requests and correspondence to: Dr. RobinChoudhury, Department of Cardiovascular Medicine,John Radcliffe Hospital, Oxford OX3 9DU, United
of �70% luminal stenosis. Plaque rupture was Kingdom. E-mail: [email protected].
2
2
2
2
2
2
2
R E F E R E N C E S
1. Rothwell PM. The high cost of notfunding stroke research: a comparisonwith heart disease and cancer. Lancet2001;357:1612–16.
2. Coull AJ, Rothwell PM. Early risk ofrecurrence by subtype of ischemicstroke in population-based incidencestudies. Neurology 2004;62:569–73.
3. Rothwell PM, Eliasziw M, Gut-nikov SA, Warlow CP, BarnettHJM (CET Collaboration). Endar-terectomy for symptomatic carotidstenosis in relation to clinical sub-groups and timing of surgery. Lancet2004;363:915–24.
4. Rothwell PM, Eliasziw M, Gut-nikov SA, et al. Analysis of pooleddata from the randomised controlledtrials of endarterectomy for symptom-atic carotid stenosis. Lancet 2003;361:107–16.
5. Rothwell PM, Gutnikov SA, WarlowCP, Collaboration EC. Reanalysis ofthe final results of the European Ca-rotid Surgery Trial. Stroke 2003;34:514–23.
6. Barnett HJ, Taylor DW, Eliasziw M,et al. Benefit of carotid endarterec-tomy in patients with symptomaticmoderate or severe stenosis. N EnglJ Med 1998;339:1415.
7. Wardlaw J. Carotid imaging for sec-ondary stroke prevention in routinepractice. Int J Stroke 2008;3:20–32.
8. Streifler J, Eliasziw M, Fox A, et al.Angiographic detection of carotidplaque ulceration. Comparison withsurgical observations in a multicenterstudy. North American SymptomaticCarotid Endarterectomy Trial. Stroke1994;25:1130–2.
9. Wasserman BA, Wityk RJ, TroutHH, Virmani R. Low-grade carotidstenosis: looking beyond the lumenwith MRI. Stroke 2005;36:2504–13.
10. Yuan C, Mitsumori LM, FergusonMS, et al. In vivo accuracy of multi-spectral magnetic resonance imagingfor identifying lipid-rich necrotic
advanced human carotid plaques. Cir-culation 2001;104:2051–6.
1. Hatsukami TS, Ross R, Polissar NL,Yuan C. Visualization of fibrous capthickness and rupture in human ath-erosclerotic carotid plaque in vivowith high-resolution magnetic reso-nance imaging. Circulation 2000;102:959 – 64.
2. Cai J-M, Hatsukami TS, FergusonMS, Small R, Polissar NL, Yuan C.Classification of human carotid ath-erosclerotic lesions with in vivo mul-ticontrast magnetic resonance imag-ing. Circulation 2002;106:1368 –73.
3. Choudhury R, Fuster V, Badimon J,Fisher E, Fayad Z. MRI and charac-terization of atherosclerotic plaque:emerging applications and molecularimaging. Arterioscler Thromb 2002;22:1065–74.
4. Yuan C, Beach KW, Smith LH, Hat-sukami TS. Measurement of athero-sclerotic carotid plaque size in vivousing high resolution magnetic reso-nance imaging. Circulation 1998;98:2666–71.
5. Yuan C. Identification of fibrous caprupture with magnetic resonance im-aging is highly associated with recenttransient ischemic attack or stroke.Circulation 2002;105:181–5.
6. Saam T, Ferguson MS, Yarnykh VL,et al. Quantitative evaluation of ca-rotid plaque composition by in vivoMRI. Stroke 2005;25:234–239.
7. Fayad ZA, Fuster V. The humanhigh-risk plaque and its detection bymagnetic resonance imaging. Am JCardiol 2001;88:42E–5E.
8. Takaya N, Yuan C, Chu B, et al.Association between carotid plaquecharacteristics and subsequent ischemiccerebrovascular events: a prospective as-sessment with MRI—initial results.Stroke 2006;37:818–23.
9. Altaf N, Daniels L, Morgan PS, et al.Cerebral white matter hyperintenselesions are associated with unstablecarotid plaques. Eur J Vasc EndovascSurgery 2006;31:8–13.
Morgan PS, et al.laque hemorrhage
by magnetic resonance imaging insymptomatic patients with mild tomoderate carotid stenosis predicts re-current neurological events. J VascSurg 2008;47:332–42.
1. Murphy RE, Moody AR, Morgan PS,et al. Prevalence of complicated ca-rotid atheroma as detected by mag-netic resonance direct thrombus imag-ing in patients with suspected carotidartery stenosis and previous acute ce-rebral ischemia. Circulation 2003;107:3053–8.
2. Parmar JP, Rogers WJ, Mugler JP, etal. Magnetic resonance imaging ofcarotid atherosclerotic plaque in clin-ically suspected acute transient isch-emic attack and acute ischemic stroke.Circulation 2010;122:2031–8.
3. Sadat U, Weerakkody RA, BowdenDJ, et al. Utility of high resolutionMR imaging to assess carotid plaquemorphology: a comparison of acutesymptomatic, recently symptomaticand asymptomatic patients with ca-rotid artery disease. Atherosclerosis2009;207:434–9.
4. Ay H, Oliveira-Filho J, Buonanno FS,et al. ‘Footprints’ of transient isch-emic attacks: a diffusion-weightedMRI study. Cerebrovasc Dis 2002;14:177– 86.
5. Brant-Zawadzki M, Atkinson D,Detrick M, Bradley WG, ScidmoreG. Fluid-attenuated inversion recov-ery (FLAIR) for assessment of cere-bral infarction. Initial clinical expe-rience in 50 patients. Stroke 1996;27:1187–91.
6. Jackson CE, Shirodaria CC, Lee JM,et al. Reproducibility and accuracy ofautomated measurement for dynamicarterial lumen area by cardiovascularmagnetic resonance. Int J CardiovascImaging 2009;25:797–808.
7. The World Health OrganizationMONICA project (monitoring trendsand determinants in cardiovasculardisease): a major international collab-oration. WHO MONICA projectprincipal investigators. J Clin Epide-
miol 1988;41:105–14.
2
3
3
3
3
3
c
J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 5 , N O . 4 , 2 0 1 2
A P R I L 2 0 1 2 : 3 8 8 – 9 6
Lindsay et al.
Acute 3-T MRI of Carotid Plaque in Stroke
396
28. Redgrave JN, Lovett JK, Gallagher PJ,Rothwell PM. Histological assess-ment of 526 symptomatic carotidplaques in relation to the nature andtiming of ischemic symptoms: the Ox-ford Plaque Study. Circulation 2006;113:2320–8.
9. Eliasziw M, Kennedy J, Hill MD,Buchan AM, Barnett HJ, GroupNASCET. Early risk of stroke after atransient ischemic attack in patientswith internal carotid artery disease.CMAJ 2004;170:1105–9.
0. Fisher M, Paganini-Hill A, Martin A,et al. Carotid plaque pathology:thrombosis, ulceration, and strokepathogenesis. Stroke 2005;36:253–7.
1. Biasi G, Froio A, Diethrich E, et al.Carotid plaque echolucency increases
the risk of stroke in carotid stenting:the Imaging in Carotid Angioplastyand Risk of Stroke (ICAROS) Study.Circulation 2004;110:756–62.
2. Hong M-K, Mintz G, Lee C, et al.Serial intravascular ultrasound evi-dence of both plaque stabilizationand lesion progression in patientswith ruptured coronary plaques: ef-fects of statin therapy on rupturedcoronary plaque. Atherosclerosis2007;191:107–14.
3. Fuster V, Badimon JJ, Badimon L.Clinical-pathological correlations ofcoronary disease progression and regres-sion. Circulation 1992;86:III1–11.
4. Geroulakos G, Domjan J, NicolaidesA, et al. Ultrasonic carotid arteryplaque structure and the risk of ce-rebral infarction on computed to-
mography. J Vasc Surg 1994;20:263– 6. r35. Tegos TJ, Sabetai MM, NicolaidesAN, et al. Correlates of embolicevents detected by means of trans-cranial doppler in patients with ca-rotid atheroma. J Vasc Surg 2001;33:131– 8.
36. Cai J, Hatsukami TS, Ferguson MS,et al. In vivo quantitative measure-ment of intact fibrous cap and lipid-rich necrotic core size in atheroscle-rotic carotid plaque: Comparison ofhigh-resolution, contrast-enhancedmagnetic resonance imaging andhistology. Circulation 2005;112:3437– 44.
Key Words: atherosclerosis yarotid arteries y magnetic
esonance imaging y stroke.
