Characterization of Carotid Plaque Hemorrhage
A CT Angiography and MR Intraplaque Hemorrhage Study
Background and Purpose— The main objective of this study was to evaluate CT angiographic (CTA) features that are able to predict the presence of intraplaque hemorrhage (IPH) as defined by MR-IPH.
Methods— One hundred sixty-seven consecutive patients (mean age 69 years, SD 12.8; 58 females) underwent both MR-IPH and CTA within 3 weeks. MR-IPH, the gold standard, was performed at 1.5 T using a neurovascular phased-array coil as a coronal T1-weighted 3-dimensional fat-suppressed acquisition. CTA was performed using a 4-slice or a 64-slice CT machine and evaluated, blinded to MR-IPH findings, for carotid stenosis, plaque density, and plaque ulceration. Plaque density was defined as the mean attenuation of plaque at the site of maximum stenosis and 2 sections above and below. Plaque ulceration was defined as outpouching of contrast into the plaque at least 2 mm deep on any single plane.
Results— Prevalence of IPH increased at higher degrees of carotid stenosis. Mean CT plaque density was higher for plaques with MRI-defined IPH (47 Hounsfield units) compared with without IPH (43 Hounsfield units; P=0.02). However, significant overlap between distributions of plaque densities limited the value of mean plaque density for prediction of IPH. CTA plaque ulceration had high sensitivity (80.0% to 91.4%), specificity (93.0% to 92.3%), positive predictive value (72.0% to 71.8%), and negative predictive value (95.0% to 97.9%) for prediction of IPH. Interobserver agreement for presence/absence of CTA plaque ulceration was excellent (κ=0.80).
Conclusions— CTA plaque ulceration, but not mean CTA plaque density, was useful for prediction of IPH as defined by the MR-IPH technique.
The benefits of surgery in recently symptomatic patients with severe carotid stenosis have been clearly demonstrated.1–3 In clinical practice, decision-making with regard to carotid revascularization still relies primarily on severity of luminal narrowing as determined by ultrasound and/or other angiographic modalities. There are, however, subgroups of patients such as asymptomatic patients or symptomatic patients with moderate stenosis in which the benefits of revascularization over medical therapy are less clearcut.1,4 In such patients, decision-making based on stenosis alone can be more difficult and there is a need for more advanced methods of risk stratification to guide optimal therapy.
Histopathologic studies have shown that certain morphological characteristics of the carotid atherosclerotic plaque such as intraplaque hemorrhage (IPH), ruptured or ulcerated fibrous cap, or large necrotic lipid core are associated with increased risk of stroke, findings independent of the severity of luminal stenosis.5,6 From a histopathologic point of view, neoangiogenesis is closely associated with plaque progression and is likely the primary source of IPH at sites of microvessel incompetence.7 Angiogenic factors contribute to the proliferation of vasa vasorum, the formation of immature vessels, and loss of basement membrane around functional capillaries. This process initiates leakage of red blood cells into the plaque and induces a cycle of inflammation and neovascularization, which increases the risk of plaque rupture and plaque ulceration.7
There is, thus, currently significant research interest in developing imaging modalities that are able to define these vulnerable features in vivo and complement luminal stenosis assessment. Multiple research groups have shown the robustness of MRI to qualitatively and quantitatively define carotid plaque morphology with respect to histology of excised carotid specimens as the gold standard.8–13 MRI has been shown to be able to classify carotid plaques according to the histopathologic classification of the American Heart Association and, moreover, the presence of vulnerable features, as depicted by in vivo MRI, has been shown to be associated with increased risk of subsequent ischemic events on clinical follow-up.14,15
MR-IPH is a previously described MRI technique consisting of a 3-dimensional T1-weighted spoiled gradient-echo sequence that is heavily fat-suppressed and therefore very sensitive to the presence of hemorrhagic products within the carotid plaque.11,16,17 It is a robust technique that can be easily performed on commercial MR machines and highlights carotid IPH as high signal against a suppressed background (Figure 1). Previous work has shown MR-IPH to have high diagnostic accuracy for detection of carotid IPH with sensitivity ranging from 84% to 100% and specificity from 80% to 88% with histology as the gold standard.11,18 The reproducibility of the technique, including inter- and intraobserver variability, has also been shown to be excellent.11 In symptomatic patients, the presence of IPH, as defined by MR-IPH, was significantly greater in the patients’ ipsilateral vessels compared with the contralateral, asymptomatic side (60% versus 36%, χ2 P<0.001), particularly for vessels of only moderate stenosis.16 The presence of IPH on MRI has also been shown to be associated with an increased risk of subsequent neurological events in both asymptomatic and symptomatic patients with mild to moderate carotid stenosis.17
In parallel with these new developments in MRI of carotid plaque, there is little doubt that, with multislice technology, CT angiography (CTA) has fast emerged as 1 of the noninvasive modalities of choice for carotid stenosis due to its high spatial resolution, speed, and ready availability.19 Compared with MRI, there has, however, been relatively less work focusing on CT identification of the vulnerable plaque. This is not surprising due to the inherent superior soft tissue contrast of MRI, but several recent studies suggest that CTA may provide some information on carotid plaque morphology.20–23 The main objective of this imaging study was, therefore, to identify CTA features that may predict presence of carotid IPH as defined by the MR-IPH technique, which was used as the gold standard method.
Materials and Methods
This study had approval from the local ethics committee (Study No. 411-2004); requirement for written informed consent was waived for this retrospective study.
At our institution, MR-IPH is routinely performed as an additional sequence as part of our clinical MRI protocol for evaluation of stroke/transient ischemic attack. This protocol also includes routine brain imaging (diffusion-weighted, fluid-attenuation inversion recovery, and T1-weighted sequences) as well as contrast-enhanced MR angiography from the aortic arch to the circle of Willis. From a prospectively collected database, we identified all patients who underwent both MR-IPH and CTA, within a 3-week interval, for evaluation of carotid stenosis, during a period ranging from December 2003 to December 2008. Patient referral for the clinical stroke MRI stroke protocol (which included contrast MR angiography) and/or CTA was at the discretion of the referring physicians but at our institution, decision for revascularization is often based on both MR angiography and CTA, because the combination of noninvasive tests has been shown to be more accurate locally.
MR-IPH was performed on a 1.5-T GE Twin Speed MR machine (GE Medical Systems, Milwaukee, Wis) using an 8-channel neurovascular phased-array coil (USA Instruments) as a free-breathing coronal T1-weighted magnetization-prepared 3-dimensional gradient-echo acquisition (TR 6.7 ms, TE 1.7 ms, TI 20 ms, flip angle 15° with 2-mm thickness, field of view 350×300 mm, effective pixel size 0.94×0.94×1 mm [interpolated], number of excitations=3). The sequence included a selective water-excitation radiofrequency pulse (Spectral Inversion at Lipids, SPECIAL; GE Healthcare) to suppress fat and the effective inversion time (TI 20 ms) was chosen to null the blood signal. The resulting acquisition time was 3.5 minutes.
Assessment of the MR-IPH studies involved reading of coronal source data together with standard multiplanar reformats in the axial and sagittal plane on commercial Picture Archiving Communications System (PACS) workstations (AGFA Impax Version 4.5). MR-IPH was reviewed by 1 of 4 neuroradiologists as to presence or absence of IPH blinded to CTA findings. A positive scan was diagnosed if high signal material (at least twice the signal intensity of the adjacent sternocleidomastoid muscle) was seen within the wall of the carotid artery at the site of the stenosis. The presence or absence of high signal was recorded for both carotid bifurcations for each patient. MR-IPH was used as the gold standard technique in this study.
All CTA examinations from December 2003 through September 2005 were performed using a LightSpeed Plus 4-section CT scanner (GE Healthcare). Images were obtained from C6 to the vertex by using the helical high-speed mode with 7.5 mm/rotation and 1.25×1.25-mm collimation (120 kVp, 350 mA). All subsequent examinations, from October 2005 through December 2008, were performed by using a LightSpeed VCT 64-section CT scanner (GE Healthcare). Images on the 64-section CT scanner were obtained from the aortic arch through the vertex at a thickness of 0.625 mm (120 kVp, auto-mA). Intravenous access was through an antecubital vein by using an 18- or 20-gauge angiocatheter. A total of 100 to 120 mL iohexol (Omnipaque 300; GE Healthcare) or iodixanol (Visipaque 320; GE Healthcare) was injected at a rate of 4.0 to 5.0 mL/s with a 17-second delay or the use of SmartPrep software (GE Healthcare) at the pulmonary artery.
Postprocessing multiplanar reformats were created at the CT operator’s console. Coronal and sagittal multiplanar reformat images were created 7.0 mm thick spaced by 3 mm. All images were viewed on Impax 4.5 (AGFA Healthcare, Mortsel, Belgium) PACS workstations. Images were initially reviewed on CTA settings (window 96, level 150 Hounsfield units [HU]), which were then modified as required on an individual basis to better depict the residual stenotic internal carotid artery lumen, the vessel wall, and plaque and also to decrease beam-hardening artifact from dense calcifications. CTA was evaluated, blinded to MR-IPH findings, for (1) severity of stenosis; (2) plaque density; and (3) presence or absence of plaque ulceration. Both carotid vessels were evaluated for each patient.
Stenosis was measured using the axial source data and multiplanar reformats according to standard North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria.2,24 Plaque density was measured by placing a circular region of interest of approximately 75% of the plaque diameter in noncalcified portions of the plaque and recording the average HU (Figure 2). The mean plaque density for the plaque was defined as the average of the HU of the plaque at the site of maximum stenosis and at 2 CT sections above and below, that is, a total of 5 HU measurements for each lesion.
Plaque ulceration was defined as contrast outpouching into the plaque at least 2 mm deep on any single plane (Figure 1). Presence or absence of plaque ulceration was recorded after review of CTA axial source images with sagittal and coronal multiplanar reformats by 2 independent neuroradiologists; for the purposes of this study, surface irregularity without definite ulceration was recorded as absence of ulceration because the evidence regarding their clinical significance is unclear.
All data were analyzed by using the statistical software package SPSS for Windows (Version 12.0.0; SPSS, Chicago, Ill). Comparison between means was performed using the Mann-Whitney U test for nonparametric data. A P value of <0.05 was defined as statistically significant.
Receiver operating characteristic curve analysis was used to determine the most appropriate plaque density cutoff value to classify presence or absence of plaque hemorrhage as defined by MR-IPH. Plaques were classified as low density (0 to 50 HU) and mixed density (51 to 120 HU) based on their mean CT attenuation values.23,25
Sensitivity, specificity, positive predictive values, and negative predictive values, with 95% CIs, were calculated for the presence of CT plaque ulceration to predict the presence or absence of IPH as defined by MRI based on standard 2×2 contingency tables. Interobserver agreement was calculated using Cohen κ statistics with 0 to 0.40, 0.40 to 0.60, 0.60 to 0.80, and 0.80 to 1 defined as poor, moderate, good, and excellent agreement, respectively.
One hundred sixty-seven consecutive patients (mean age 69 years, SD 12.8; 58 females) underwent both MR-IPH and CTA during the time period of the study. One hundred fifty-three patients were symptomatic with suspected strokes or transient ischemic attacks and the remainder was asymptomatic. Of a total of 334 arteries analyzed, 15 arteries were excluded due to occlusions (n=13) or previous carotid endarterectomy (n=2), yielding 319 arteries for analysis. MR-IPH showed 56 cases of carotid IPH.
Stenosis Severity and IPH
Distribution of carotid stenosis was as follows: mild stenosis: n=193 (60.5%), moderate stenosis: n=60 (18.8%), and severe stenosis: n=66 (20.7%). Mean stenosis severity for the group with positive MR-IPH studies was 58% (SD 25) compared with 20% (SD 26) for the group with negative MR-IPH studies (P=0.001, Mann-Whitney U test). Figure 3 shows the distribution of cases of hemorrhagic plaques (ie, positive MR-IPH studies) per category of stenosis severity. This confirms that plaque hemorrhage was more frequent at higher degrees of stenosis severity, although there was a nonnegligible number of cases of IPH seen at mild and moderate degrees of stenosis.
Mean CT Plaque Density and IPH
Mean CT plaque density was higher for plaques with MRI-defined IPH (47 HU, SD 15) compared with without IPH (43 HU, SD 14; P=0.02, Mann-Whitney U test). However, significant overlap between distributions of plaque densities suggests that mean plaque density is of limited value in distinguishing between the 2 groups. This is confirmed by the poor receiver operating characteristics of CT mean plaque density for classification of IPH as shown in Figure 4.
Two hundred twenty-three plaques (69.9%) were of low density on CT (mean ≤50 HU) compared with 96 plaques (30.1%) of mixed density (51 to 120 HU). Thirty-six (64.3%) of the cases of plaque hemorrhage were seen in plaques of low CT density compared with 20 cases (35.7%) of plaque hemorrhage in plaques with mixed CT density. The presence of low density on CTA would have sensitivity of 64.3% (95% CI 51 to 76), specificity of 28.9% (95% CI 24 to 35), positive predictive value of 16% (95% CI 12 to 22), and negative predictive value of 79% (95% CI 70 to 86).
CT Plaque Ulceration and IPH
CT plaque ulceration had excellent diagnostic accuracy for presence or absence of MRI-defined IPH. The 2×2 contingency tables for presence/absence of CTA plaque ulceration and presence/absence of IPH on MRI for the 2 CTA readers is shown in the Table.
Diagnostic performance for Reader 1 was as follows: sensitivity of 80% (95% CI 68 to 88), specificity of 93% (95% CI 89 to 96), positive predictive value of 72% (95% CI 59 to 81), and negative predictive value of 95% (95% CI 92 to 98).
Diagnostic performance for Reader 2 was as follows: sensitivity of 91.4% (95% CI 81 to 96), specificity of 92.3% (95% CI 89 to 95), positive predictive value of 71.8% (95% CI 60 to 81), and negative predictive value of 97.9% (95% CI 95 to 99).
Interobserver agreement between 2 readers was excellent (κ=0.80).
Our imaging study is the first to show a strong in vivo association between CTA plaque ulceration and IPH as defined by MR-IPH. Given the increasing use of CTA to quantify stenosis, these results emphasize the importance of diagnosing plaque ulceration on CTA and recognizing their importance as a significant marker of stroke risk.
Plaque ulceration has long been recognized as a risk factor for neurological complications independent of stenosis. A histopathologic study comparing microscopic plaque morphology from patients with and without stroke symptoms involving patients from the Asymptomatic Carotid Atherosclerosis Study (ACAS) and NASCET studies showed that plaque ulceration was significantly more common in plaques taken from symptomatic patients than those without symptoms (36% versus 14%; P<0.001).26 Moreover, in patients with high-grade stenosis in the NASCET study, the presence of ulceration on conventional digital subtraction angiography increased the relative risks of stroke from 1.24 to 3.43.27 Furthermore, a large study comparing surface morphology with detailed histology showed that ulceration on digital subtraction angiography was strongly associated with both plaque rupture and IPH.28 Complementing these findings, our results confirm that CTA plaque ulceration can be used a surrogate marker for IPH and therefore as a surrogate risk factor for stroke.
Although the sensitivity of single-slice CTA for detection of plaque ulceration has previously been questioned,29 it is likely that with the advent of multislice technology and the improved higher spatial resolution, CTA will now show much better diagnostic performance.20 For instance, recent studies have shown high sensitivity and specificity of multidetector row CTA (up to 93.75% and 98.59%, respectively) for detection of plaque ulceration with respect to histology.30 Both Wintermark et al and de Weert et al showed that in vivo multidetector row CTA performed well for detection of plaque ulceration.20,31 Their results are of similar order to ours.
In our experience, a significant advantage of CTA compared with conventional digital subtraction angiography is the ability to reformat the studies in any orthogonal plane without any significant loss of spatial resolution. We found that the axial source images were particularly useful for detecting plaque ulceration, especially when calcification tended to obscure the plaque in the sagittal or coronal planes (Figure 5). This experience has been mirrored by others; for instance, axial source images have been shown by Saba et al to have the highest sensitivity and specificity (90.9% and 87.9%, respectively) for detection of plaque ulceration.32
We did not find mean plaque density to be a useful factor for prediction of IPH, as defined by MRI, in our study. There was significant overlap between the mean plaque densities between the hemorrhagic and the nonhemorrhagic plaque groups and receiver operating characteristic analysis showed that there was no suitable cutoff value that would allow prediction of IPH with any reasonable degree of sensitivity and specificity. This is in concordance with the results of several previous in vivo and in vitro studies. Wintermark et al compared identified different plaque components such as fibrous tissue, hemorrhage, lipid core, and calcification on histological carotid endarterectomy specimens, and by directly matching CT images with the histological sections in an unblinded manner, calculated the range of HU for each component. The mean in vivo CT HU was 97.5 (95% CI 53.5 to 141.6), 46.4 (95% CI 6.6 to 86.2), 32.6 (95% CI −7.4 to 72.5), and 256.7 (95% CI 216.3 to 297.1) for hemorrhage, fibrous tissue, lipid core, and calcification, respectively. They also found significant overlap between the HU of different plaque components. In the second part of the study, they used the derived fibrous values to calibrate an automatic classifier to predict plaque components on CT images based on HU. Not unsurprisingly, CTA did not perform well in classifying plaque hemorrhage or lipids because of the overlap with fibrous tissue; the performance of CTA, however, improved slightly for detection of large plaque hemorrhage.
Similarly, de Weert et al characterized plaque components on histology and calculated HU of plaque components by matching corresponding CT and histological sections. The measured HU were 88 (SD 18), 25 (SD 19), and 657 (SD 16) for fibrous tissue, lipid core, and calcification, respectively. They calculated cutoff values of 60 HU for differentiation of lipid core and fibrous tissue and 130 HU for differentiation of fibrous tissue and calcification based on receiver operating characteristic analysis. They, however, found that calcification was a significant confounding factor because of beam-hardening artifacts and that quantification of the lipid core was only accurate in mildly calcified plaques. However, de Weert et al also found that hemorrhage and thrombus could not readily be identified from fibrous tissue and lipid core due to the overlap in plaque densities of these components; these data corroborate our findings.
Walker et al, using similar methodology to our study, compared plaque density on single-slice CTA to histology of excised carotid endarterectomy specimens.29 They found similar limitations regarding the use of CTA for plaque components. They found that the poor reliability of HU measurements for the prediction of the amount of lipid, fibrous tissue, or hemorrhagic components within an individual plaque may be explained, at least partly, by the great heterogeneity observed on histological examination of individual plaques. Walker et al argue that the main reason behind the poor performance of CTA was that the relatively homogeneous appearance of plaque on CT imaging did not adequately represent the plaque heterogeneity that was evident on microscopic analysis.29 We agree with Walker et al and accept this factor as 1 of the limitations of our study. Our approach to calculate mean density of the plaque on CTA images does not take into account the heterogeneity of the plaque and its different constituents and the mean HU values calculated is the overall mean HU value of the plaque. Because plaques have a homogeneous appearance on CTA, it is not possible to target a specific component such as hemorrhage with the naked eye and specifically place a region of interest over a particular component. More complex approaches such as the automatic classifier developed by Wintermark et al will be therefore necessary if CTA is to become useful for plaque characterization.20 However, these approaches are still very preliminary and not robust enough for clinical practice currently. It was the objective of our current study to test a simple, rapid, and practical method, easily performed within minutes, on commercially used workstations for plaque density measurements rather than the more complex approaches as used by Wintermark et al.
Finally, it is interesting that few studies have shown that plaques with low density on CTA were often found to have hemorrhagic components on histology.23,33 However, the low density seen on CTA is clearly not directly measuring the CT attenuation of hemorrhage but reflecting the presence of large necrotic lipid cores within the plaque. Although lipid cores and hemorrhage can be associated in vulnerable plaques, 1 feature is often seen without the other and it is possible that this association was found to be significant in these studies mainly because of the small sample size of the population studied. In our study, 36 of the cases of plaque hemorrhage were seen in plaques with low CT density (≤50 HU) compared with 20 cases of plaque hemorrhage in plaques with mixed density (50 to 120 HU). Despite moderate sensitivity and negative predictive value, the presence of low density on CTA had very poor specificity and negative predictive value for presence or absence of IPH.
One of the limitations of our study is that we used MR-IPH as the gold standard technique for detection of IPH. Histology of excised carotid specimens remains the gold standard, but this was not available in this study. However, histology is an ex vivo standard with its own limitations such as plaque damage during surgical excision. However, MRI is currently, in our opinion, the best in vivo modality for IPH and the technique used in this study has previously been validated in comparison with histology.11 High signal on T1 within the plaque has good sensitivity and specificity for IPH; previous studies and our own observations during this study have shown that high signal is not related to the presence of plaque calcium, a recognized but very unusual cause of high T1 signal on MRI.18 We did not perform additional high-resolution MRI to fully characterize plaque morphology in this study and limited ourselves to the identification of IPH. This would have required additional MRI sequences with additional weightings, requiring significantly more imaging time (≥30 minutes). These additional sequences would not have been possible currently at our institution in a busy clinical setting and in such a large number of patients. This puts into context the practicality and robustness of our technique with an acquisition time of 3 minutes only and which can easily be added to any clinical MRI stroke imaging protocol.
Finally, it was outside of the scope of this imaging-based study to correlate the imaging findings with clinical events. There are, however, multiple studies from several groups correlating positive MR-IPH studies with increased risk of stroke.14,17,34 More recently, Singh et al showed that the presence of MR-defined IPH predicted the risks of subsequent neurological complications in a cohort of asymptomatic men.35
In our study, presence of CT plaque ulceration, but not mean CT plaque density, was useful for prediction of IPH as defined by our gold standard technique, MR-IPH. Recent studies suggest that CTA may have high accuracy in detection of plaque ulceration and we found high interobserver agreement for CT plaque ulceration.30 Given the increasing use of CTA to quantify carotid stenosis, this study emphasizes the importance of diagnosing plaque ulceration on CTA and recognizing their importance as a marker for plaque hemorrhage and thus as a potential surrogate marker for stroke risk.
- Received January 17, 2010.
- Accepted February 9, 2010.
Barnett HJ, Taylor DW, Eliasziw M, Fox AJ, Ferguson GG, Haynes RB, Rankin RN, Clagett GP, Hachinski VC, Sackett DL, Thorpe KE, Meldrum HE. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med. 1998; 339: 1415–1425.
Redgrave JN, Lovett JK, Gallagher PJ, Rothwell PM. Histological assessment of 526 symptomatic carotid plaques in relation to the nature and timing of ischemic symptoms: the Oxford Plaque Study. Circulation. 2006; 113: 2320–2328.
Virmani R, Kolodgie FD, Burke AP, Finn AV, Gold HK, Tulenko TN, Wrenn SP, Narula J. Atherosclerotic plaque progression and vulnerability to rupture: angiogenesis as a source of intraplaque hemorrhage. Arterioscler Thromb Vasc Biol. 2005; 25: 2054–2061.
Saam T, Ferguson MS, Yarnykh VL, Takaya N, Xu D, Polissar NL, Hatsukami TS, Yuan C. Quantitative evaluation of carotid plaque composition by in vivo MRI. Arterioscler Thromb Vasc Biol. 2005; 25: 234–239.
Hatsukami TS, Ross R, Polissar NL, Yuan C. Visualization of fibrous cap thickness and rupture in human atherosclerotic carotid plaque in vivo with high-resolution magnetic resonance imaging. Circulation. 2000; 102: 959–964.
Moody AR, Murphy RE, Morgan PS, Martel AL, Delay GS, Allder S, MacSweeney ST, Tennant WG, Gladman J, Lowe J, Hunt BJ. Characterization of complicated carotid plaque with magnetic resonance direct thrombus imaging in patients with cerebral ischemia. Circulation. 2003; 107: 3047–3052.
Coombs BD, Rapp JH, Ursell PC, Reilly LM, Saloner D. Structure of plaque at carotid bifurcation: high-resolution MRI with histological correlation. Stroke. 2001; 32: 2516–2521.
Takaya N, Yuan C, Chu B, Saam T, Underhill H, Cai J, Tran N, Polissar NL, Isaac C, Ferguson MS, Garden GA, Cramer SC, Maravilla KR, Hashimoto B, Hatsukami TS. Association between carotid plaque characteristics and subsequent ischemic cerebrovascular events: a prospective assessment with MRI—initial results. Stroke. 2006; 37: 818–823.
Cai JM, Hatsukami TS, Ferguson MS, Small R, Polissar NL, Yuan C. Classification of human carotid atherosclerotic lesions with in vivo multicontrast magnetic resonance imaging. Circulation. 2002; 106: 1368–1373.
Murphy RE, Moody AR, Morgan PS, Martel AL, Delay GS, Allder S, MacSweeney ST, Tennant WG, Gladman J, Lowe J, Hunt BJ. Prevalence of complicated carotid atheroma as detected by magnetic resonance direct thrombus imaging in patients with suspected carotid artery stenosis and previous acute cerebral ischemia. Circulation. 2003; 107: 3053–3058.
Altaf N, MacSweeney ST, Gladman J, Auer DP. Carotid intraplaque hemorrhage predicts recurrent symptoms in patients with high-grade carotid stenosis. Stroke. 2007; 38: 1633–1635.
Wintermark M, Jawadi SS, Rapp JH, Tihan T, Tong E, Glidden DV, Abedin S, Schaeffer S, Acevedo-Bolton G, Boudignon B, Orwoll B, Pan X, Saloner D. High-resolution CT imaging of carotid artery atherosclerotic plaques. AJNR Am J Neuroradiol. 2008; 29: 875–882.
de Weert TT, Ouhlous M, Meijering E, Zondervan PE, Hendriks JM, van Sambeek MR, Dippel DW, van der Lugt A. In vivo characterization and quantification of atherosclerotic carotid plaque components with multidetector computed tomography and histopathological correlation. Arterioscler Thromb Vasc Biol. 2006; 26: 2366–2372.
Bartlett ES, Walters TD, Symons SP, Fox AJ. Quantification of carotid stenosis on CT angiography. AJNR Am J Neuroradiol. 2006; 27: 13–19.
Fisher M, Paganini-Hill A, Martin A, Cosgrove M, Toole JF, Barnett HJ, Norris J. Carotid plaque pathology: thrombosis, ulceration, and stroke pathogenesis. Stroke. 2005; 36: 253–257.
Eliasziw M, Streifler JY, Fox AJ, Hachinski VC, Ferguson GG, Barnett HJ. Significance of plaque ulceration in symptomatic patients with high-grade carotid stenosis. North American Symptomatic Carotid Endarterectomy Trial. Stroke. 1994; 25: 304–308.
Lovett JK, Gallagher PJ, Hands LJ, Walton J, Rothwell PM. Histological correlates of carotid plaque surface morphology on lumen contrast imaging. Circulation. 2004; 110: 2190–2197.
Walker LJ, Ismail A, McMeekin W, Lambert D, Mendelow AD, Birchall D. Computed tomography angiography for the evaluation of carotid atherosclerotic plaque: correlation with histopathology of endarterectomy specimens. Stroke. 2002; 33: 977–981.
Saba L, Caddeo G, Sanfilippo R, Montisci R, Mallarini G. CT and ultrasound in the study of ulcerated carotid plaque compared with surgical results: potentialities and advantages of multidetector row CT angiography. AJNR Am J Neuroradiol. 2007; 28: 1061–1066.
de Weert TT, Cretier S, Groen HC, Homburg P, Cakir H, Wentzel JJ, Dippel DW, van der Lugt A. Atherosclerotic plaque surface morphology in the carotid bifurcation assessed with multidetector computed tomography angiography. Stroke. 2009; 40: 1334–1340.
Saba L, Caddeo G, Sanfilippo R, Montisci R, Mallarini G. Efficacy and sensitivity of axial scans and different reconstruction methods in the study of the ulcerated carotid plaque using multidetector-row CT angiography: comparison with surgical results. AJNR Am J Neuroradiol. 2007; 28: 716–723.
Oliver TB, Lammie GA, Wright AR, Wardlaw J, Patel SG, Peek R, Ruckley CV, Collie DA. Atherosclerotic plaque at the carotid bifurcation: CT angiographic appearance with histopathologic correlation. AJNR Am J Neuroradiol. 1999; 20: 897–901.
U-King-Im JM, Tang TY, Patterson A, Graves MJ, Howarth S, Li ZY, Trivedi R, Bowden D, Kirkpatrick PJ, Gaunt ME, Warburton EA, Antoun NM, Gillard JH. Characterisation of carotid atheroma in symptomatic and asymptomatic patients using high resolution MRI. J Neurol Neurosurg Psychiatry. 2008; 79: 905–912.