| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Medical Imaging, University of Toronto, Toronto, Ontario,
Canada.
2 Department of Medical Imaging, Sunnybrook Health Sciences Centre, 2075 Bayview
Ave., AG46, Toronto, Ontario, Canada M4N 3M5.
3 Institute for Clinical Evaluative Sciences (ICES), Sunnybrook Health Sciences
Centre, Toronto, Ontario, Canada.
4 Department of Neurology and Regional Stroke Centre, Sunnybrook Health Sciences
Centre, Toronto, Ontario, Canada.
5 Division of Vascular Surgery, Sunnybrook Health Sciences Centre, Toronto,
Ontario, Canada.
Received September 1, 2005;
accepted after revision October 25, 2005.
Richard Bitar is the recipient of a Canadian Heads of Academic Radiology
(CHAR) resident grant and a Canadian Institutes of Health Research (CIHR)
fellowship.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Patients referred for investigation of cerebrovascular disease by MRI were enrolled. Reasons for referral included transient ischemic attack/amaurosis fugax, acute infarct, remote infarct, or asymptomatic carotid disease. Of the 348 patients initially scanned, 17 were excluded from the analysis. The final patient population included 331 patients (199 men, 132 women; mean age, 67.7 years). Patients were scanned using MRDTI, a 3D, T1-weighted, fat-suppressed spoiled gradient echo that exploits the T1 shortening effects of methemoglobin, directly visualizing hemorrhage/thrombus in the vessel wall, thus identifying complicated plaque. Complicated plaque was defined as a high signal within the atherosclerotic plaque at least twice the signal intensity of muscle.
RESULTS. Forty-three of 331 patients (13%) had complicated upper thoracic aortic atherosclerotic disease, arch vessel atherosclerotic disease, or both. The upper thoracic aorta was involved in 36 of 43 patients (83.7%), and the left subclavian artery was involved in 14 of 43 patients (32.6%). Both the right subclavian artery and the brachiocephalic artery were involved in one of 43 patients (2.3%). Complicated carotid plaque was seen in 25 of 43 patients (58.1%).
CONCLUSION. MRDTI can be applied in the detection of complicated plaque in the upper thoracic aorta and arch vessels. Complicated plaque was identified in 13% of the patient population. The upper thoracic aorta was the most common site involved. This technique could be useful for the screening of asymptomatic at-risk patients.
Keywords: cardiovascular imaging • hemorrhage • MR arteriography • MRI • peripheral vascular disease
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Atherosclerotic plaques that lead to morbidity and mortality are usually modestly stenotic, often not seen by angiography [6]. The AHA has developed a classification system for atherosclerotic plaques [7], with intraplaque hemorrhage and thrombosis as markers that define atherosclerotic plaques as complicated (AHA type VIb/c) and at an increased risk of causing symptoms [7]. A study by Kolodgie et al. [8] confirmed the association in the coronary arteries between intraplaque hemorrhage and plaque instability.
As hemoglobin within intraplaque hemorrhage matures, it goes through various states, one of which, methemoglobin, causes shortening of T1 relaxation and results in the high signal intensity seen in T1-weighted imaging [9]. By exploiting the T1 shortening effects of methemoglobin, MR direct thrombus imaging (MRDTI) can directly visualize intraplaque hemorrhage and therefore complicated atherosclerotic plaques when directly compared with histology [10, 11]. MRDTI-detected complicated plaque is prevalent in symptomatic and, to a lesser extent, in asymptomatic carotid arteries [12]. Detection of methemoglobin within complicated plaques should be feasible in other vascular beds, raising the prospect of surveying the vascular system to identify high-risk vessels before they become symptomatic.
The purpose of this study was to apply MRDTI to the carotids, upper thoracic aorta, and arch vessels to assess the feasibility of detecting AHA type VIb/c plaque as part of systemic atherosclerotic disease in patients referred for evaluation of cerebrovascular disease. We wanted to compare the prevalence and location of complicated plaque with the known prevalence of stenoocclusive atherosclerotic disease in the upper thoracic aorta and arch vessels.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patients
From September 1, 2003, to February 28, 2005, patients referred for
investigation of cerebrovascular disease by MRI were enrolled. Reasons for
referral included transient ischemic attack (TIA)/amaurosis fugax, acute
infarct, remote infarct, or asymptomatic carotid disease.
MR Direct Thrombus Imaging Sequence
Patients were scanned using a 1.5-T GE Twin Speed MR scanner (GE
Healthcare) and an eight-channel neurovascular phased-array coil (USA
Instruments). The MRDTI sequence used a free-breathing 3D T1-weighted
fat-suppressed spoiled gradient echo (TR/TE, 6.7/1.7; flip angle, 15°)
with 2-mm thickness; field of view, 300 mm2; matrix size,
3202; effective pixel size, 0.94 mm x 0.94 mm x 1 mm
(interpolated); and 3 averages. Fat suppression was achieved using SPECIAL
(Spectral Inversion at Lipids), a GE Healthcare proprietary technique.
Scanning time was 04:13 minutes.
The field of view included the upper thoracic aorta (defined in this study as the aortic arch and the upper portion of the descending thoracic aorta); the arch vessels (the brachiocephalic and the right and left subclavian arteries); and the carotid arteries.
Contrast-Enhanced MR Angiography
Contrast-enhanced MR angiography (CEMRA) was performed as part of the
clinical imaging protocol. The sequence was a 3D spoiled gradient echo (TR/TE,
4.2/1.1; flip angle, 40°) with 3-mm thickness; field of view, 300
mm2; matrix size, 320 x 224; effective pixel size, 0.9 mm
x 1.3 mm x 1.5 mm (interpolated); and 1 average. An IV contrast
agent (Omniscan Egadodiamide, 0.1 mmol/kg, Nycomed Amersham) was injected at a
rate of 2 mL/s. Stenosis levels were classified as wall irregularities, less
than 50% stenosis, greater than 50% stenosis, or occluded artery.
Image Analysis
Images were reviewed by a radiologist with a specific interest in vascular
MRI. For the MRDTI sequence, analysis involved examination of the carotid
system for complicated plaque, defined as high signal within the
atherosclerotic plaque at least twice the signal intensity of muscle
[11], followed by the upper
portions of the thoracic aorta and the arch vessels. The location of
complicated plaque was recorded and then categorized as located in the upper
thoracic aorta (pre-left subclavian artery, post-left subclavian artery, or
both); the brachiocephalic artery; or the right subclavian, left subclavian
artery, or both. CEMRA was qualitatively assessed for vessel occlusion,
stenosis, vessel wall irregularities, or all three.
Statistical Analysis
A multivariate logistic regression analysis was used. Odds ratios and their
95% confidence intervals (CIs) and p values were reported for all
significant findings (SAS version 9.1).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Prevalence and Distribution of Complicated Plaque
Of the 331 patients, 43 (13%) had complicated upper thoracic aorta plaque,
arch vessel plaque, or both. The upper thoracic aorta was involved in 36 of
331 patients (10.9%) (Fig. 1).
Aortic involvement was mostly distal to the left subclavian (36/43 patients,
83.7%), three patients (3/36, 8.3%) had both pre- and post-left subclavian
disease, and three patients (3/36, 8.3%) had combined complicated left
subclavian and upper thoracic aorta atherosclerotic plaque.
|
|
|
|
|
|
Patients without complicated carotid plaque—Of the 242 patients with no complicated carotid plaque, 18 had upper thoracic aorta and/or arch vessel disease (18/242, 7.4%). Fifteen of the 18 patients (83.3%) were symptomatic for cerebrovascular disease (TIA/amaurosis fugax, infarct). Ten of these patients (10/18, 55.6%) had complicated upper thoracic plaque but distal to the left subclavian artery; therefore, potential emboli from this plaque could not account for their symptoms (Fig. 4). The remaining five patients all had symptoms consistent with an anterior circulation TIA. Only one of these patients (1/5, 20%) had complicated plaque proximal to the left subclavian artery; therefore, this complicated proximal aortic plaque could have accounted for this patient's symptomatology. The others had complicated plaque in the left subclavian artery.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Application of MRDTI to identify complicated plaque in the thoracic aorta and arch vessels represents a novel use of this sequence previously shown to identify complicated plaque in the carotid arteries when compared with histology [10-12]. To our knowledge, no studies have directly visualized complicated plaque in vivo using MRI in the arch vessels, and only one other study has reported in vivo visualization of AHA type VIb/c plaque in the descending thoracic aorta [13].
Using this technique, we have reported the prevalence of complicated atherosclerotic plaque in both the carotid arteries and the upper thoracic aorta and arch vessels. Previous studies using digital subtraction angiography have shown that subclavian and carotid artery disease commonly coexist [14]. In this study, 25 of 89 patients (28.1%) with complicated carotid disease had complicated aortic plaque, arch vessel plaque, or both, and of those, 11 patients (11/89, 12.4%) had complicated subclavian artery atherosclerotic disease. Five of these 11 patients (5/89, 5.6%) had carotid stenosis greater than 70% and were potential surgical candidates. This finding has clinical implications because perioperative stroke is increased in patients undergoing carotid endarterectomy if severe subclavian artery disease is present related to the compromise of the posterior circulation [15].
MRDTI is a simple, quick, noninvasive method for imaging complicated atherosclerotic plaque, easily added to standard MRI investigations of patients with cerebrovascular disease. As a 3D sequence, MRDTI inherently has a better signal-to-noise ratio than the 2D sequences commonly used in MRI studies of atherosclerotic disease [13, 16, 17]. The combination of the high signal generated by methemoglobin and suppression of the flowing blood signal by repeated radiofrequency pulses provides excellent contrast to identify complicated plaque. This aids in the detection of minimally or nonstenosing plaque. Because 3D data are acquired, multiplanar reformation allows visualization of complicated structures such as the aortic arch. Despite the 3D acquisition, few studies were degraded by physiologic motion (8/348, 2.3%).
MRDTI evaluation of the upper thoracic aorta and arch vessels provides early detection of complicated but minimally stenosing asymptomatic disease. Recent studies have suggested that asymptomatic and symptomatic peripheral arterial disease appear to have equivalent ischemic risk [18, 19]. The Prevention of Atherothrombotic Disease Network has published a "call to action" [20] that includes the recommendation to initiate a screening program for detecting asymptomatic peripheral arterial disease in high-risk patients. MRDTI is able to detect complicated plaque and may allow identification of high-risk disease in at-risk asymptomatic patients.
Transesophageal echocardiography (TEE) historically was the technique of choice to image the thoracic aorta [19, 21]. One of the first in vivo human studies evaluated atherosclerotic plaques in the descending thoracic aorta [13], in which the accuracy of MRI versus TEE to identify the composition, size, and extent of different types of atherosclerotic plaques was compared in 10 patients. A total of 25 thoracic aortic plaques were identified, and three were defined as complicated plaques with hemorrhage/thrombus (AHA type VIb/c). The authors attributed this low number of complicated plaques to their technique, which may not have been sensitive enough to detect this type of complicated plaque [13]. Our ability to detect complicated plaque might be explained by the optimized parameters exploiting the T1-shortening effects of methemoglobin, the benefits of 3D over 2D acquisition, and the use of a gradient echo sequence, previously shown to be better at detecting intraplaque hemorrhage when compared with fast spin-echo sequences [17].
Other studies have also used MRI for detection of asymptomatic atherosclerotic disease of the human aorta. In 102 patients undergoing coronary angiography for suspected coronary artery disease, Taniguchi et al. [16] detected descending thoracic atherosclerotic plaque in 61% of their patients. Jaffer et al. [22] used MRI to detect subclinical aortic atherosclerosis in 318 patients from the offspring cohort of the Framingham Heart Study. Forty percent of their patients had subclinical aortic atherosclerosis, with 6.3% located in the thoracic aorta and 38.8% in the abdominal aorta. Neither of these studies evaluated the type of plaque. The difference in prevalence of thoracic aortic plaque reported in these two studies is high, ranging from 6.3% to 61%. Our reported prevalence lies within this range (36/331, 10.9%), although the total burden of atherosclerotic disease is likely higher because we only imaged AHA type VIb/c plaque. These variations in prevalence could also be attributed to the different patient populations in the studies.
Some previous studies have used MRI to visualize vessel wall disease directly; other studies have used MR angiography (MRA) to detect stenoocclusive atherosclerosis. In a study of 100 patients with known or clinically suspected cervical vascular disease undergoing 3D gadolinium-enhanced MRA, Ersoy et al. [15] found stenotic disease of the subclavian artery in 57% of their patients, with a predilection for the left side (68%).
Atherosclerotic disease was also found in the brachiocephalic, carotid, and vertebral arteries. In the study reported here, 32.6% (14/43) of the patients with complicated upper thoracic aorta disease, arch vessel disease, or both had complicated atherosclerotic disease in the subclavian arteries, also with a predilection for the left side (13/14, 92.9%). The lower prevalence again likely reflects the detection of AHA type VIb/c disease only.
The presence of combined aortic and carotid atherosclerotic disease has been previously shown using TEE [23, 24]. In a study of 89 patients with acute ischemic strokes, Guo et al. [23] evaluated the distribution of aortic arch and intra/extracranial cerebral atherosclerosis. They found that the incidence of carotid plaques was significantly higher among patients who had aortic arch atherosclerosis compared with those who did not (71.2% vs 21.6%). In our study, we found that patients with complicated plaque in the upper thoracic aorta, the arch vessels, or both also had more complicated carotid plaque compared with those who did not (58% [25/43] vs 22.2% [64/288]) (Fig. 4B). Demopoulos et al. [24] evaluated 45 patients with carotid stenosis greater than or equal to 50% and stroke or TIAs within 6 weeks, matching these patients for age, sex, and hypertension with controls. They found protruding aortic arch atheromas in 38% of the patients with carotid disease. In our study, we also found that patients with complicated plaque in the carotid arteries were more likely to have complicated plaque in the upper thoracic aorta, arch vessels, or both (28.1% [25/89]). We have found the prevalence of AHA type VIb/c plaque is approximately 30-60% of the reported total atherosclerotic burden from the previous studies, representing the unrecognized high-risk lesions not characterized by alternative techniques.
Thirty-six patients had complicated upper thoracic aorta plaque, 91.7% of which (33/36 patients) were distal to the left subclavian artery, the region where type B aortic dissections originate. An association between atherosclerotic plaques and the site of thoracic aortic dissection has previously only been shown in a small number of patients [25]. However, AHA type VIb/c atherosclerotic plaque could perhaps represent the earliest stage of intramural hematoma (IMH), which could subsequently develop into a dissection. IMH is a variant of aortic dissection that can result either from bleeding of the vasa vasorum in the media or from a penetrating atherosclerotic ulcer causing hematoma formation in the aortic wall [26, 27]. We do not believe that the vessel wall high signal seen in these patients represents advanced IMH, which typically is associated with sudden onset and severe excruciating chest or back pain [25, 28, 29].
Mural thrombosis is commonly found on the surface of advanced atherosclerotic lesions and aneurysms, especially in the aorta [30-32]. Its incidence and pathogenesis remains unclear, but it has been identified in nonaneurysmal aortas, although the reported incidence is extremely low (0.45%) [33]. Differentiation between intraplaque hemorrhage and atherosclerotic surface thrombosis is problematic [11, 33, 34] because it is difficult to localize the high signal seen in the vessel wall at this resolution [34, 35]. However, both of these represent subtypes of AHA type VI disease [8] and are detected by MRDTI.
Our study does have some limitations. Correlation of our technique with other imaging such as TEE may have been useful to better delineate the exact location of the high signal seen in the vessel walls. The coverage of the aorta was limited to the aortic arch and the upper descending thoracic aorta, so atherosclerotic disease in the aortic root and lower descending thoracic aorta was not fully evaluated, potentially missing important atherosclerotic disease. Follow-up of patients with type VIb/c atherosclerotic disease would provide information regarding the clinical relevance of detecting such disease (to see if they developed complications from their high-risk disease). Also, some of the high signal seen by MRDTI may be false-positive results; however, the MRDTI technique has been previously shown to have a low rate of false-positivity, with a positive predictive value of 93% [11].
The MRDTI technique should benefit from respiratory and pulsatile motion compensation using methods such as breath-holding, respiratory gating, ECG gating, and navigator pulses. This should improve evaluation of complicated plaque in the aortic root/ascending thoracic aorta, important because these aortic segments have been implicated in coronary artery disease and cerebral thromboembolism [36, 37]. Higher spatial resolution should also allow better delineation of the location of the high signal, allowing differentiation among intraplaque hemorrhage, intramural hematoma, and mural thrombosis [35]. This could have clinical significance, as recent a report by Ganaha et al. [38] suggests that intramural hematoma caused by a penetrating atherosclerotic ulcer has a worse clinical prognosis than one caused by bleeding from the vasa vasorum. Natural history studies would define the role of this technique in the screening of high-risk patients with asymptomatic disease.
In conclusion, MRDTI is a quick and simple technique to detect complicated plaque in the upper thoracic aorta and arch vessels. Complicated plaque was identified in 13% of the patient population in this study, and the upper thoracic aorta was the most common site involved. This technique could be useful for screening asymptomatic at-risk patients. Identification of these at-risk patients before they become symptomatic could be important because it will allow opportunities for medical intervention to halt and potentially reverse the disease process.
Acknowledgments
We thank the MRI technologists at our institution.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
