Relationship between carotid plaque surface morphology and perfusion: a 3D DCE-MRI study
- 527 Downloads
This study aims to explore the relationship between plaque surface morphology and neovascularization using a high temporal and spatial resolution 4D contrast-enhanced MRI/MRA sequence.
Materials and methods
Twenty one patients with either recent symptoms or a carotid artery stenosis ≥40% were recruited in this study. Plaque surface morphology and luminal stenosis were determined from the arterial phase MRA images. Carotid neovascularization was evaluated by a previously validated pharmacokinetic (PK) modeling approach. K trans (transfer constant) and v p (partial plasma volume) were calculated in both the adventitia and plaque.
Image acquisition and analysis was successfully performed in 28 arteries. Mean luminal stenosis was 44% (range 11–82%). Both adventitial and plaque K trans in ulcerated/irregular plaques were significantly higher than smooth plaques (0.079 ± 0.018 vs. 0.064 ± 0.011 min−1, p = 0.02; 0.065 ± 0.013 vs. 0.055 ± 0.010 min−1, p = 0.03, respectively). Positive correlations between adventitial K trans and v p against stenosis were observed (r = 0.44, p = 0.02; r = 0.55, p = 0.01, respectively).
This study demonstrates the feasibility of using a single sequence to acquire both high resolution 4D CE-MRA and DCE-MRI to evaluate both plaque surface morphology and function. The results demonstrate significant relationships between lumen surface morphology and neovascularization.
KeywordsDynamic contrast enhanced MRI (DCE-MRI) Contrast enhanced MR angiography (CE-MRA) Carotid atherosclerotic plaque Plaque ulceration Neovascularization
Stroke is one of the leading causes of death and disability worldwide  with ischemic strokes accounting for 90% of the total reported cases . Twenty percent of ischemic strokes are attributed to carotid atherosclerosis . Carotid luminal stenosis measured by ultrasound plays an important role in screening high risk patients and is the standard criteria for surgical referral . Whilst only patients with severe stenosis (≥70%) have shown a benefit from carotid endarterectomy (CEA), patients with moderate stenosis (<70%) have only shown marginal beneficial effects from surgical treatment [5, 6]. There is an increasing recognition that other factors, such as plaque surface morphology, may have higher predictive value in moderately stenosed plaques .
The appearance on X-ray angiography of plaque surface irregularity, or ulceration, has been correlated with increased risk of ischemic stroke . This correlation applies for both moderate and severe stenosis . Carotid ulceration is found to be more common in patients with downstream micro emboli . Ulceration has also been associated with high risk factors including plaque rupture, intraplaque hemorrhage and a large lipid-rich necrotic core (LRNC) , indicating this is a sensitive marker of plaque instability  and risk of stroke [8, 11, 12]. Noninvasive imaging modalities offer high accuracy for ulcer detection as well as stenosis measurement [13, 14, 15]. Among different imaging modalities, contrast-enhanced MR angiography provides high resolution imaging allowing visualization of plaque ulceration [13, 16, 17].
In this study we have applied a 4D sequence  which acquires data suitable for contrast enhanced (CE) MRA and dynamic contrast enhanced (DCE)-MRI processing. DCE-MRI has been demonstrated as a non-invasive method to quantify the extent of neovasculature within carotid plaques with high accuracy and reproducibility [19, 20]. Previous studies using DCE-MRI have shown that plaque neovascularization is closely associated with high risk features, such as inflammation , and is related to ischemic events . In this study, we investigated the correlation between plaque surface morphology and neovascularization.
Materials and methods
The study protocol was reviewed and approved by the local ethics committee and written informed consent was obtained. Twenty one patients who had a carotid stenosis of at least 40% on Duplex ultrasound using NASCET criteria  were recruited in total. Ten were symptomatic (nine males, mean age 72 years, range 59–86 years) with a recent history (less than 6 months) of transient ischemic attack (TIA), and 11 were asymptomatic (seven males, mean age 78 years, range 68–87 years).
Summary of imaging parameters for the multi-contrast MRI protocol at 3T
Pre/post-contrast BB T1w 3D FSE
Flip angle (°)
Variable flip angle
FOV (mm × mm × mm)
140 × 140 × 64
140 × 140 × 67
160 × 160 × 66
140 × 140 × 62
Acquired pixel size (mm × mm × mm)
0.5 × 0.5 × 2.0
0.6 × 0.6 × 1.4
1.0 × 1.0 × 1.0
0.6 × 0.6 × 1.4
Reconstruction pixel size (mm × mm × mm)
0.5 × 0.5 × 2.0
0.3 × 0.3 × 0.7
0.3 × 0.3 × 1.0
0.3 × 0.3 × 0.7
Receiver bandwidth (±kHz)
1 min 35 s
2 × 6 min 26 s
4 min 42 s
6 min 23sa
The T1w and DCE images were reformatted into the axial plane with a 0.7 mm slice thickness using a dedicated workstation (AW4.6, GE Healthcare, Buc, France). For each patient, the carotid arteries on both sides were analyzed. Carotid vessel wall and lumen boundaries at each slice were manually drawn according to the T1w images, using a DICOM viewer (OsiriX 5.5.2, Pixmeo, Geneva, Switzerland). Plaque was defined as a focal wall thickness ≥1.5 mm . The subtracted CE-MRA image at the fifth phase, which had the highest image contrast in the carotid artery, was used for morphological analysis. Plaque surface morphology was classified as either ulcerated/irregular or smooth using the maximum intensity projection (MIP) of CE-MRA images, based on previous published standards , by two experienced reviewers, both having more than 2 years of experience in carotid imaging. The two reviewers were blinded to the patients clinical information and made their judgement independently. The classification result from the first reviewer was used for further analysis. Luminal stenosis was measured for each plaque on both CE-MRA source image and 3D pre-contrast T1w images separately according to the North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria [23, 27]. The stenosis measured from CE-MRA source image was used for further analysis. This method has a reported high intra- and inter-observer reproducibility . IPH was defined as bright pixels on the DTI image.
The statistical analysis was performed using R (version 3.2.2). The Shapiro–Wilk’s test was used to test normality assumptions. The Kappa statistic was performed to evaluate interrater agreement in ulceration detection. The Student’s t test was used to compare the PK parameters for smooth versus ulcerated/irregular plaques defined by the first reviewer, and also to compare hemorrhage and non-hemorrhage plaques. Pearson’s correlation was used to assess the relationship between PK parameters and luminal stenosis. The concordance correlation coefficient (CCC) was reported to assess the relationship between the stenosis measured by CE-MRA and 3D T1w images. The CCC and Bland–Altman plots were used to compare the differences between linear and non-linear assumptions of image signal and Gd concentration. The inter-rater agreement in PK parameters was assessed using the intraclass correlation coefficient (ICC). A Chi-squared tests was used to compare proportions of patients with IPH. Parametric distributions were presented as mean ± SD. p values <0.05 were considered statistically significant.
Of the 21 patients, three patients did not finish the examination due to discomfort, four carotid arteries were excluded from data analysis due to motion artefacts and four arteries were excluded due to occlusion. The rest of the images were of diagnostic quality, and used for further analysis. Atherosclerotic plaques were identified in the remaining 28 arteries, with a mean luminal stenosis of 44% (range 11–82%). Fourteen plaques were classified as ulcerated/irregular and fourteen were classified as smooth. In the 28 arteries analyzed, two independent reviewers agreed in 26 (93%) cases for plaque surface morphology classification (\(\kappa = 0.79\); 95% CI 0.56–1.00). The PK parameters (K trans and v p in adventitia and plaque) were normally distributed (p > 0.05). For the luminal stenosis, the correlation between CE-MRA and 3D T1w was good (r = 0.89).
IPH was found in eight plaques. Within this group, six arteries had ulcerated plaques and two had smooth plaques. In the 20 plaques without IPH, we noted that smooth plaques were more common (12 had smooth plaques), however, these respective proportions were not significantly different (p = 0.28). Plaques with hemorrhage have higher mean adventitial K trans values compared with plaques with no hemorrhage (0.078 ± 0.019 vs. 0.069 ± 0.015 min−1), however the difference here was not significant (p = 0.2).
The agreement for the four measured DCE parameters between the two reviewers was good. The ICC for adventitial K trans, plaque K trans, adventitial v p and plaque v p are 0.88, 0.93, 0.81 and 0.71, respectively.
To the best of our knowledge, this is the first application of a 4D sequence to acquire both high spatial and temporal resolution DCE-MRI and CE-MRA of the carotid artery, allowing for combined morphological and functional assessment of carotid atherosclerosis from a single contrast administration. The results show that adventitial and plaque K trans were higher in ulcerated/irregular plaques compared to smooth plaques. Adventitial K trans and v p were also positively correlated with the degree of luminal stenosis.
Carotid plaque ulceration, characterized as an intimal defect with the intraplaque component exposed to the lumen , can form emboli and thrombosis leading to ischemic neurologic symptoms. Sitzer et al. showed plaque ulceration and lumen thrombus are the main sources of cerebral microemboli . Multiple studies have also shown that plaque ulceration is associated with cerebral infarcts, TIA and amaurosis fugax [8, 11]. In this study, plaque enhancement could be observed in both smooth and ulcerated plaques and the derived VVI showed significant differences in PK parameters between both of these plaque types. The results demonstrate that ulcerated plaques are more prevalent to neovascularization and possibly inflammation. Plaque ulceration or fibrous cap discontinuity may induce more inflammatory cell infiltration into the plaque, which in turn further weakens the fibrous cap and eventually causes the plaque to rupture. This suggests quantitative analysis of DCE-MRI could provide more information than conventional contrast-enhanced morphological images in identifying lesions at higher risk.
While only adventitial K trans and v p are correlated with luminal stenosis, no significant correlation was observed between whole plaque PK values and the stenosis. This might be due to a lower sensitivity as a result of averaging PK parameters across the whole plaque region. While various plaque components such as IPH, lipid core and fibrous tissue have different PK values , simple averaging of PK parameters across the plaque region may be insufficient for quantitative analysis and risk stratification. A positive correlation between luminal stenosis and adventitial K trans and v p was observed in this study (Fig. 4a, b), indicating a relationship between plaque growth and adventitial neovascularization. These microvessels within the atherosclerotic lesion have shown close association with inflammatory cells  which indicate that these could serve as a pathway for the macrophage entry into the plaque that leads to chronic inflammation hence promoting plaque development.
The presence of intraplaque hemorrhage (IPH) has also been found to be associated with increased adventitial K trans . In our study, plaques with IPH had higher adventitial K trans, however the difference was not significance. The reason for this may due to lack of statistical power, as only a small number of plaques (eight) contained IPH. Although the exact etiology of the IPH is still unclear it is suggested that most likely the rupture of fragile and leaky neovessels that vascularize the developed plaque leads to IPH .The histopathological evidence also supports the relationship of IPH to the presence of neovessels . In our study, IPH was absent in most of the plaques. The correlation between PK parameters and surface morphology also applies to the plaques without IPH, indicating ulceration and a high degree of stenosis were two other factors, apart from IPH, which could present with increased plaque neovascularization and inflammation.
The use of multi-contrast MRI allows for detailed analysis of plaque composition and vulnerability assessment, which could provide more insight than simple luminal stenosis measurements . This study demonstrates that a combined 4D DCE-MRI/MRA acquisition in combination with multi-contrast 3D sequences can aid in the identification of high-risk plaque features. Our study also demonstrates that these features can be obtained in a clinically acceptable scan time of approximately 30 min.
There are several limitations in this study. Firstly, no histological validation was performed on the plaques. However the correlation between DCE-MRI and neovascularization as well as inflammation has been demonstrated in previous studies [19, 21, 28], and CE-MRA is considered reliable for observing plaque surface morphology [13, 16]. The second limitation is that only a small patient cohort was included in this study. As discussed above, the K trans difference between IPH groups was not significant, this might be due to the resultant lack of statistical power. Thirdly, the view-ordering used in this study could smooth the temporal information, especially during the injection period. However, this is not considered to be a significant effect and other studies have used similar analysis for various DCE applications, including breast [38, 39], prostate  and also animal models . The last limitation is that a thick coronal imaging slice (1.4 mm) was used in this study. This might bring errors in calculating the PK values for the thin vessel wall. Future protocol design could consider increasing axial resolution for a more precise analysis.
This study investigated the relationship between PK modeling with carotid plaque surface morphology by simultaneously acquiring 4D DCE-MRI and CE-MRA. Adventitial and plaque K trans from DCE-MRI analysis is significantly higher in ulcerated plaques than smooth ones. Also, adventitial K trans and v p are correlated with luminal stenosis.
The Vasa Vasorum Image tool was kindly provided by Dr. William Kerwin of the Vascular Imaging Laboratory of the University of Washington. We thank Dr. Andrew Gill for insightful discussions.
The project was supported by the Addenbrooke’s Charitable Trust and the NIHR comprehensive Biomedical Research Centre award to Cambridge University Hospitals NHS Foundation Trust in partnership with the University of Cambridge.
JY participated in the study design, contributed to data acquisition, analysis, interpretation, and drafted the manuscript. GCM was involved in study design, funding application, recruiting the patients and revising the manuscript. AJP and MJG designed the study, supervised the data acquisition, analysis, interpretation and prepared the manuscript. AU was involved in patient recruitment, data analysis and revising the manuscript. TD was involved in data analysis and revising the manuscript. ANP developed and tested the direct thrombus imaging sequence, and revised the manuscript. PD was involved in patient scanning and revising the manuscript. SH was involved in patient recruitment, scanning and revising the manuscript. ZT was involved in discussion and revising the manuscript. JHG supervised the study, and revised the manuscript.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
- 2.Goldstein LB, Bushnell CD, Adams RJ, Appel LJ, Braun LT, Chaturvedi S, Creager MA, Culebras A, Eckel RH, Hart RG (2011) Guidelines for the primary prevention of stroke a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 42(2):517–584CrossRefPubMedGoogle Scholar
- 15.U-King-Im J, Trivedi R, Cross J, Higgins N, Graves M, Kirkpatrick P, Antoun N, Gillard J (2004) Conventional digital subtraction X-ray angiography versus magnetic resonance angiography in the evaluation of carotid disease: patient satisfaction and preferences. Clin Radiol 59(4):358–363CrossRefPubMedGoogle Scholar
- 23.Barnett H, Taylor D, Eliasziw M, Fox A, Ferguson G, Haynes R, Rankin R, Clagett G, Hachinski V, Sackett D (1998) North American Symptomatic Carotid Endarterectomy Trial Collaborators. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. N Engl J Med 339(20):1415–1425CrossRefPubMedGoogle Scholar
- 24.Zhu C (2014) High resolution black blood magnetic resonance imaging of atherosclerotic plaque. PhD Thesis, University of CambridgeGoogle Scholar
- 25.Priest AN, Joubert I, Hilborne S, Hunter S, Bowden DJ, Graves MJ, Baglin T, Gillard JH, John D (2013) MR Direct Thrombus Imaging with optimised signal and improved lipid suppression. In: Proceedings of the 21st scientific meeting, International Society for Magnetic Resonance in Medicine, Salt Lake City, p 1280Google Scholar
- 39.Bedair R, Graves MJ, Patterson AJ, McLean MA, Manavaki R, Wallace T, Reid S, Mendichovszky I, Griffiths J, Gilbert FJ (2015) Effect of radiofrequency transmit field correction on quantitative dynamic contrast-enhanced MR imaging of the breast at 3.0 T. Radiology 279(2):368–377CrossRefPubMedGoogle Scholar
- 40.Rödel S, Blaut S, Dürig E, Haroske G, Steinbach F, Kittner T, Dresden D (2014) Prostate: multiparametric magnetic resonance imaging (mpMRI) and MR-guided biopsy (MR-GB). ECR 2014/C-2079. doi: 10.1594/ecr2014/C-2079
- 41.Haeck J, Bol K, Bison S, Tiel S, Koelewijn S, Jong M, Veenland J, Bernsen M (2015) Optimized time-resolved imaging of contrast kinetics (TRICKS) in dynamic contrast-enhanced MRI after peptide receptor radionuclide therapy in small animal tumor models. Contrast Media Mol Imaging 10(6):413–420CrossRefPubMedGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.