Imaging intracranial vessel wall pathology with magnetic resonance imaging: current prospects and future directions.

Abstract

HomeCirculationVol. 130, No. 2Imaging Intracranial Vessel Wall Pathology With Magnetic Resonance Imaging Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBImaging Intracranial Vessel Wall Pathology With Magnetic Resonance ImagingCurrent Prospects and Future Directions Nikki Dieleman, MSc, Anja G. van der Kolk, MD, Jaco J.M. Zwanenburg, PhD, Anita A. Harteveld, MSc, Geert J. Biessels, MD, PhD, Peter R. Luijten, PhD and Jeroen Hendrikse, MD, PhD Nikki DielemanNikki Dieleman From the Department of Radiology (N.D., A.G.v.d.K., J.J.M.Z., A.A.H., P.R.L., J.H.), Image Sciences Institute (J.J.M.Z.), and Department of Neurology (G.J.B.), University Medical Center Utrecht, Utrecht, the Netherlands. Search for more papers by this author , Anja G. van der KolkAnja G. van der Kolk From the Department of Radiology (N.D., A.G.v.d.K., J.J.M.Z., A.A.H., P.R.L., J.H.), Image Sciences Institute (J.J.M.Z.), and Department of Neurology (G.J.B.), University Medical Center Utrecht, Utrecht, the Netherlands. Search for more papers by this author , Jaco J.M. ZwanenburgJaco J.M. Zwanenburg From the Department of Radiology (N.D., A.G.v.d.K., J.J.M.Z., A.A.H., P.R.L., J.H.), Image Sciences Institute (J.J.M.Z.), and Department of Neurology (G.J.B.), University Medical Center Utrecht, Utrecht, the Netherlands. Search for more papers by this author , Anita A. HarteveldAnita A. Harteveld From the Department of Radiology (N.D., A.G.v.d.K., J.J.M.Z., A.A.H., P.R.L., J.H.), Image Sciences Institute (J.J.M.Z.), and Department of Neurology (G.J.B.), University Medical Center Utrecht, Utrecht, the Netherlands. Search for more papers by this author , Geert J. BiesselsGeert J. Biessels From the Department of Radiology (N.D., A.G.v.d.K., J.J.M.Z., A.A.H., P.R.L., J.H.), Image Sciences Institute (J.J.M.Z.), and Department of Neurology (G.J.B.), University Medical Center Utrecht, Utrecht, the Netherlands. Search for more papers by this author , Peter R. LuijtenPeter R. Luijten From the Department of Radiology (N.D., A.G.v.d.K., J.J.M.Z., A.A.H., P.R.L., J.H.), Image Sciences Institute (J.J.M.Z.), and Department of Neurology (G.J.B.), University Medical Center Utrecht, Utrecht, the Netherlands. Search for more papers by this author and Jeroen HendrikseJeroen Hendrikse From the Department of Radiology (N.D., A.G.v.d.K., J.J.M.Z., A.A.H., P.R.L., J.H.), Image Sciences Institute (J.J.M.Z.), and Department of Neurology (G.J.B.), University Medical Center Utrecht, Utrecht, the Netherlands. Search for more papers by this author Originally published8 Jul 2014https://doi.org/10.1161/CIRCULATIONAHA.113.006919Circulation. 2014;130:192–201To date, the probable cause of ischemic stroke is often inferred from the size and location of the infarct, in combination with an evaluation of the heart and the presence of extracranial arterial occlusion or high-grade stenosis.1 Currently used conventional lumenography-based methods such as digital subtraction angiography, computed tomography angiography, and magnetic resonance (MR) angiography are used to determine the presence of such an acute occlusion or high-grade arterial stenosis. From extracranial studies, it is known that luminal narrowing may be absent in patients with severe atherosclerosis owing to arterial remodeling.2–4 Therefore, these methods do not provide information about the underlying pathological processes, which most often involve the vessel wall.5 Vessel wall changes such as vessel wall thickening, enhancement, or the presence of vulnerable atherosclerotic plaques without luminal stenosis are therefore often missed but might be of importance for a better understanding of ischemic stroke.6 Furthermore, intracranial atherosclerosis is an important cause of ischemic stroke7 and often involves the vessel wall. Patients with intracranial atherosclerosis have high recurrent stroke rates,8 and increasingly more attention is being directed to the assessment of the intracranial vessel wall, necessitating an imaging technique directly assessing the intracranial vessel wall. MR imaging (MRI) seems the most promising technique to reliably image intracranial vessel wall pathologies because of its superior soft tissue contrast. Recent advances in MRI9 have made it possible to obtain information about these abnormalities within the intracranial vessel wall, which provides an imaging tool to investigate the role of intracranial vessel wall abnormalities in the diagnosis of stroke.In this review, we discuss the current status of intracranial vessel wall MRI and its potential to identify different intracranial vessel wall pathologies. First, we present the state-of-the-art MRI methods to visualize the intracranial vessel wall and its pathology and we provide the technical background of these imaging methods. This part also focuses on imaging at different field strengths and on healthy vessel wall and different wall pathologies. Then, we give an overview of intracranial vessel wall abnormalities that may be expected for each ischemic stroke subtype. Finally, given the clinical needs and technical possibilities, we review the future clinical and technical directions of intracranial vessel wall imaging and provide a tentative recommendation for use in daily clinical practice.MRI of Intracranial Vessel Wall PathologyTechnical RequirementsTo successfully image the intracranial vessel wall, MRI methods have been developed and optimized that suppress the intracranial arterial blood signal and, in some cases, combine this with suppression of the cerebrospinal fluid (CSF) signal around the intracranial arteries proximally and distally of the circle of Willis.9–22 Obtaining black-blood signal for all intracranial arteries is essential for sufficient image contrast to assess the vessel wall and its pathology. The distal internal carotid arteries, distal vertebral arteries, and basilar artery have a predominant caudal-cranial course, which may be beneficial when the outflow of blood is used together with a transverse imaging plane perpendicular to the arteries proximally of the circle of Willis. The basilar artery especially has been shown to be well suited for vessel wall imaging because of its relative large size and straight course.20 On the other hand, compared with the arteries in the neck, the anterior, middle, and posterior branches of the circle of Willis all have a different course, and black-blood imaging based on outflow may be more difficult. Some of the black-blood imaging methods that have been developed for the neck may also be applied and optimized for intracranial vessel wall imaging. These black-blood techniques include double inversion recovery23 and techniques that are based on motion-sensitizing prepulses.24,25 A disadvantage might be that slow flow next to the vessel wall results in incomplete blood suppression, thereby overestimating vessel wall thickness.26For the detection of focal thickening of the intracranial vessel walls, signal suppression on the outside of intracranial vessel walls is a prerequisite. Most of the intracranial arteries are surrounded by CSF. In elderly patients, this amount of CSF surrounding the intracranial arteries increases as a result of brain atrophy.9 With MRI methods that provide dark CSF signal or suppress the CSF signal, the outer walls will become visible (Table 1). A combination of intracranial methods that suppress the blood and the CSF will result in the visualization of both the inner and outer arterial vessel wall. However, suppressing the CSF will cost signal-to-noise ratio (SNR) within the vessel wall, and there may be a time penalty for the optimization of MRI methods. Especially when these methods are combined with high-spatial-resolution imaging, scan times may become too long.49Table 1. Scan Parameters Used in Studies Assessing Intracranial Vessel Wall Pathology With MRIField Strength, TMRIAuthorsScan Duration, Minimum–Maximum, minFOV, Minimum–Maximum, mmResolution, Minimum–Maximum, mmTR/TE/TI, msPulse SequenceContrast Enhancement1.5T1-weighted imagingAoki et al272.24–12.00150×150×3.00.6×0.7×2.0600/15SE+Aoki et al28150×150×3.00.6×0.7×2.0400–600/8–15SE+Klein et al29120×120×2-30.4×0.5×2.0900/15FSE+Park et al13150×150×3.00.48×0.58×2.0n.g.n.g.−Küker et al30n.g.n.g.n.g.n.g.+Natori et al31250×190×1900.5×0.5×0.5500/18.3FSE−Vakil et al32220×240×5.00.9×0.9×5.0 (gap 1.5)663/15SE+1.5T2-weighted imagingAoki et al2811.20–12.00150×150×3.00.6×0.7×2.03000–4000/80–100(F)SE−Klein et al29,33,34120×120×2.50.4×0.5×2.0–0.4×0.4×2.53500/70FSE−Küker et al,30 Park et al13150×150×3.00.48×0.58×2.0n.g.TSE−1.5PD-weighted imagingAoki et al287.47150×150×3.00.6×0.7×2.02000–4000/14–30SE−Klein et al29,33120×120×2.50.4×0.5×2.0–0.4×0.4×2.51500–2600/15FSE+1.5Dual-echo imagesLam et al353.41–4.15200×200×3.00.8×0.8×3.03000/120FSE−Chen et al36*170×170×3.00.9×0.7×3.03000/19–130TSE−1.5Balanced fast field echoSparing et al37220×220×1.40.4×0.4×1.45.6/2.8FSE−3T1-weighted imagingChung et al142.88–11.38100×125×2.00.4×0.6×2.0700/23TSE−Chung et al38100×100×1.00.5×0.5×1.0800–100/8.4TSE+Lou et al,10 Ma et al11160×160×2.00.6×0.5×2.0800/8.6FSE+Mandell et al18220×220×2-30.4×0.4×2-32263/13/860FSE+Matouk et al39160×160×2.00.3×0.3×2.0590/10TSE+Pfefferkorn et al,19 Saam et al22200×200×2.00.4×0.4×2.0800/13TSE+Ryu et al21120×105×2.00.4×0.5×2.0581/20TSE−Shi et al4080×80×2.00.3×0.3×2.0700/14TSE−Skarpathiotakis et al,41 Swartz et al42160×220×2.00.4×0.6×2.02108/12/860FSE+Swartz et al42160×220×3.00.4×0.6×3.0450/21SE+Turan et al43128×128×1280.4×0.4×1.2458/16TSE+Turan et al43100×100×1000.4×0.4×1.22500/14/140TSE−Xu et al44,45130×130×2.00.5×0.5×2.0800/12DIR−Vergouwen et al46,47n.g.n.g.n.g.n.g.+3T2-weighted imagingChung et al143.20–6.87100×125×2.00.4×0.6×2.01800/78TSE−Chung et al38100×100×1.00.5×0.5×1.02000/100TSE−Li et al,48 Xu et al45130×130×2.00.5×0.5×2.03000/50FSE−Ma et al11160×160×2.00.6×0.5×2.03000/12.5FSE−Niizuma et al49120×120×2.00.2×0.5×2.02800/50.8FSE−Ryu et al21120×105×2.00.4×0.5×2.03000/80TSE−Shi et al40140×140×0.60.6×0.6×0.62500/67TSE−Skarpathiotakis et al,41 Swartz et al42160×220×2.00.3×0.4×2.03450/92FSE−Turan et al43104×104×1040.4×0.4×1.21500/66TSE−Vergouwen et al47n.g.n.g.n.g.n.g.−3PD-weighted imagingChung et al143.20–9.17100×125×2.00.4×0.6×2.01800/26TSE−Chung et al38100×100×1.00.5×0.5×1.02000/32TSE−Kim et al,15,16 Ryu et al21120×105×2.0–120×110×2.00.4×0.5×2.02500/30TSE−Kim et al17104×120×1.00.3×0.5×1.03000/30TSE−Kim et al17105×120×1.00.4×0.6×1.02000/40TSE−Ma et al11,12160×160×2.00.6×0.5×2.03000/12.5FSE−Qiao et al20200×166×450.5×0.5×0.52000/38TSE−Qiao et al20120×90×2.00.25×0.25×2.0200/38TSE−Shi et al4080×80×2.00.3×0.3×2.02400/17TSE−Zhu et al50160×160×2.00.6×0.6×2.04000/12.9FSE−3T1-weighted MRDTITuran et al519320×320×2240.7×0.7×0.78.7/3.8GE−7T1-weighted MPIR TSEvan der Kolk et al912220×180×130.8×0.8×0.86050/23/1770TSE+Mixed contrast T1+T2Majidi et al53*45395×119×10.40.13×0.13×0.133000/60TSE−FOV indicates field of view; FSE, fast spin echo; MPIR, magnetization preparation inversion recovery; MR, magnetic resonance; MRDTI, magnetic resonance direct thrombus imaging; n.g., not given; PD, proton density; SE, spin echo; TE, echo time; TI, inversion time; TR, repetition time; and TSE, turbo spin echo.*Specimen studies.Because the diameter of the intracranial arteries is very small, ranging from 2 to 3 mm proximally to <1 mm more distally,9,54,55 sufficient SNR and, concordant to that, submillimeter spatial resolution are required to allow the detection of vessel wall changes such as plaques, vessel wall thickening, and focal enhancement. Especially for vessel wall imaging of the arteries distal to the circle of Willis, a high spatial resolution is needed.9,42 To obtain such a high resolution within clinically reasonable scan times, scan coverage has to be reduced (Table 1). Most current intracranial vessel wall imaging methods therefore visualize only a relatively small portion of the intracranial arteries and focus on arteries at the level of the circle of Willis and just proximal to it.10–12,14–17,21,40,44,45,49,50 Smaller distal intracranial arterial vessel walls are beyond the scope of these sequences.42 However, a relatively low spatial resolution may well be sufficient to detect enhancement of a lesion or more generalized enhancement of the intracranial vasculature as a result of an increase in contrast-to-noise ratio caused by the contrast uptake (Table 1).27,28,30 Third, not only the in-plane resolution but also, in the case of 2-dimensional sequences, slice thickness is of importance. When the slice orientation is perpendicular to the artery of interest, the slice can be relatively thick. Still, when different segments with variable orientations to the intracranial arteries need to be assessed, multiple 2-dimensional slices need to be scanned, with each slice oriented perpendicular to the local vessel orientation. In this case, a 3-dimensional sequence with isotropic resolution may be superior because of its reconstruction possibilities that allow reconstruction of image planes perpendicular to the local vessel orientation afterward. The currently available MRI methods use different resolutions and slice thicknesses (Table 1), and the images are often acquired at a lower spatial resolution with the use of interpolation by zero filling during reconstruction to view the images at a higher resolution.44,45,50 Although zero filling may yield a smoother visualization of the vessel walls, the acquired resolution determines the actual available information of the vessel wall lesions.Magnetic Field StrengthTo image the thin intracranial vessel wall at a high resolution, high SNR is essential. To obtain acceptable SNR, one can move to higher field strengths because SNR scales linearly with field strength.9,56 The gain in SNR can, in principle, be used for either a shorter scan time or a higher spatial resolution compared with lower field strengths.52 Because this will allow higher spatial resolution within a reasonable scan time, intracranial vessel wall imaging is performed mainly at higher field strengths nowadays (Table 1). The first studies of intracranial vessel wall imaging were performed at a field strength of 1.5 T (Table 1). Because of the lower achievable spatial resolution of these scanners for sufficient SNR, most interest was paid to intracranial vessel wall lesions with a high contrast-to-noise ratio (Table 1) such as vividly enhancing lesions of intracranial arterial walls caused by cerebral vasculitis. With the availability of clinical 3 T MRI scanners, smaller lesions that were typically below the detection limit of 1.5 T MRI became detectable. At 3 T, intracranial vessel wall imaging showed the presence of intracranial plaques and enhancement of intracranial lesions.10–14,21,33-35,40,44,45–49 Most of these lesions were first described in the posterior circulation (distal vertebral and basilar artery),10–12,16,29,34,38 whereas more recent studies also showed these abnormalities in the middle cerebral arteries.13–15,17,21,29,31,35,36,40,44,45–49,57 In the last decade, 7 T MRI scanners have become available in a number of research centers and hospitals worldwide. With the even higher attainable SNR at 7 T, it became possible to perform isotropic vessel wall imaging with a larger coverage both in vivo9 and ex vivo53 (Table 1). This made it possible to reconstruct each artery with a different course in different orientations without the loss of detail (Figure 1).9 Pulse sequence optimization at the various field strengths may even allow more detail, for instance, characterization of vessel wall lesions and visualization of more peripherally located arteries. The developments and optimization of vessel wall imaging protocols at higher field strengths (7 and 3 T) may teach us disease-specific key imaging findings of the intracranial arterial walls at high resolution. Head-to-head comparison studies are currently not available but should be performed in the future to examine the appropriate imaging method and MRI field strength for the detection and characterization of intracranial vessel wall lesions.Download figureDownload PowerPointFigure 1. A 58-year-old man presented with transient hemiparesis of the left arm and leg caused by transient ischemic attack of the right middle cerebral artery. The 7 T 3-dimentional magnetization preparation inversion recovery turbo spin-echo transverse images before (A) and after (B) contrast administration show a focal atherosclerotic lesion of the right M1 branch just after its bifurcation and a lesion in the distal basilar artery, both enhancing after contrast administration (arrowheads). C and D, Sagittal and coronal reconstructions of the contrast-enhanced focal lesion of the right M1 branch.Healthy Vessel WallImaging vessel wall pathology has been shown to be easier than imaging the thinner, normal, healthy vessel wall. The latter clearly requires a higher resolution and good contrast both between the inner side of the arterial vessel wall and the blood and between the outer vessel wall and the CSF (Figure 2).9 In young patients with limited brain atrophy, the outer vessel wall of the more distal intracranial arteries may be located next to the brain parenchyma, and CSF suppression alone might not be sufficient to show the outer wall of these arteries.9 Visualization of the thicker vessel walls of the intracranial vessels proximal to the circle of Willis and the proximal middle cerebral artery may be easier compared with more distal, very thin vessel walls of small vessels such as the distal middle cerebral artery and the posterior and anterior cerebral arteries. In clinical practice, the healthy vessel wall was shown to be rather difficult to depict with the currently available imaging methods.9,20,21,27,42,44,45,48 However, visualization of the healthy intracranial vessel wall may be important for several reasons. First, pathological changes in the vessel wall, especially when nonfocal pathology is concerned, may be discovered only when a comparison is made between the possible pathological vessel wall and its healthy contralateral side. Second, similar to the intima-media thickness measurements of the carotid artery,58 clear visualization of the intracranial arterial walls may allow thickness measurements of these vessel walls, even when no pathological process is present. Thickness measurements of the intracranial vessel walls may serve as a general parameter or biomarker for the status of these intracranial arteries, as for the carotid artery.59 At this moment, however, this is not yet feasible; at least 2 voxels within the vessel wall are needed for dedicated measurements; therefore, an in-plane resolution of ≤0.2 mm is required. Third, clear visualization of the arterial vessel walls, together with the presence of atherosclerotic changes and focal plaques, may allow the assessment of possible arterial remodeling with a widening of the arterial diameter at the location of the plaque.4 On the other hand, the resolution and contrast-to-noise ratio required to visualize the healthy vessel wall may increase the scan time for the vessel wall imaging sequence to a degree that this method cannot be easily added to a standard clinical MRI examination. On the basis of the indication of intracranial vessel wall imaging, the methods may be optimized to focus on visualization of focal plaques and enhancement of vessel walls rather than clear visualization of the healthy vessel walls.Download figureDownload PowerPointFigure 2. A 79-year-old man with a history of transient ischemic attacks of the right hemisphere presented with dysphasia and a right-sided facial paralysis resulting from a small cortical ischemic stroke of the frontal part of the middle cerebral artery territory as seen on diffusion-weighted magnetic resonance images (not shown). A 7 T transverse 3-dimensional magnetization preparation inversion recovery turbo spin-echo image with zoomed box, before contrast administration, shows smooth, healthy vessel wall (arrows) and thickening of the left and right M1 branch (arrowheads) probably of atherosclerotic origin.Nonenhancing Vessel Wall LesionsWhen intracranial vessel wall lesions are evident, they may be visualized with intracranial vessel wall imaging rather easily, as has been shown by multiple studies.11–14,17,20,21,29,31,33–35,40,42,44,45,47–50 With intracranial vessel wall imaging, multiple lesions in multiple segments often are detected that give a picture of the total burden of intracranial atherosclerosis. Detection and characterization of individual small intracranial vascular lesions may provide additional information (Figure 2). Plaques may be characterized by their location and shape. Similar to the rest of the arterial vasculature, there also may be a preferential location of intracranial atherosclerotic lesions. Although this location needs to be established, it can be hypothesized that a preferential location at the bifurcations of the different arteries exists similar to the rest of the arterial vasculature.60 With respect to the shape of the arterial lesions, it is expected that when the resolution of intracranial vessel wall imaging is sufficient (≤0.5 mm3), different subtypes of atherosclerotic plaques can be distinguished from complete concentric (around the arterial lumen) to more focal and eccentric on 1 side of the lumen of an intracranial arterial segment.61 Still, only eccentric atherosclerotic plaques have been observed so far.42,46,47 Furthermore, the size of the lesions and irregularity might be assessed. To do so, the presence of atherosclerotic lesions may be combined with the assessment of the arterial wall thickness and possible remodeling of the arterial lumen at the side of the lesion.Vessel Wall (Lesion) EnhancementWith improvement of intracranial vessel wall imaging methods, recent studies have shown enhancement of smaller atherosclerotic plaques in intracranial vessels (Table 1).9–12,32,38,41–43,46,47,51 MRI methods for contrast enhancement typically rely on the use of T1-weighted sequences. Contrast-enhanced sequences benefit from a strong effect of the gadolinium-based contrast agents as a result of T1 shortening caused by the contrast uptake. This will result in higher signal intensity on these T1-weighted MRI sequences when contrast is taken up by the vessel wall or plaque (Figure 1). For the detection of enhancement, an additional T1-weighted MRI sequence without contrast (precontrast) is used to compare with the contrast-enhanced sequence (postcontrast). Still, the added value of both precontrast and postcontrast T1-weighted sequences for the detection of intracranial vessel wall enhancement has not yet been established. When the enhancement is very clear such as with vasculitis, a postcontrast scan may be sufficient (Figures 3 and 4).18,19,22 However, the contrast enhancement can also be modest, or intracranial vessel wall lesions may already have a relatively hyperintense signal before the administration of contrast (Figure 2).9 In these more difficult cases, precontrast and postcontrast MRI vessel wall imaging is required to establish the presence of contrast uptake. Whether there is an optimal time point between contrast administration and peak enhancement needs to be determined. Preliminary results from 7 T MRI examinations show a contrast-to-noise ratio peak after 20 minutes of contrast administration (Figure 4). However, this needs to be established in a larger group of patients and might vary between individual cases and pathologies. Finally, the exact pathophysiological mechanisms of contrast uptake in the intracranial arterial walls and atherosclerotic plaques need to be established. These mechanisms may differ on the basis of the kind of pathology (eg, vasculitis, atherosclerotic plaque) imaged. Contrast enhancement may occur with more generalized or focal inflammation of the vessel wall, together with neovascularization and endothelial contrast leakage,62 or as a result of vasa vasorum in atherosclerotic plaques. For instance, pathological enhancement of a plaque in a vessel supplying a stroke territory has been observed within 4 weeks of ischemic stroke, and the strength and presence of enhancement are closely related to the time between stroke and vessel wall assessment.41 The degree of enhancement is thought to be closely correlated with the level of inflammatory activity, presumably as a result of neovascularization and increased endothelial permeability.63Download figureDownload PowerPointFigure 3. A 63-year-old man presented with aphasia resulting from infarctions of the left and right middle cerebral artery territory and a cerebral hemorrhage in the left hemisphere as seen on diffusion-weighted magnetic resonance and fluid-attenuated inversion-recovery images (not shown) suspected of having cerebral vasculitis. A and B, Postcontrast 3 T 3-dimensional T1-weighted volumetric isotropic turbo spin-echo acquisition (sequence adjusted from Qiao et al;20 repetition time/echo time, 1500/35 milliseconds) vessel wall images on 2 different levels showing smooth, healthy vessel wall (arrows) and thickening and enhancement of the vessel wall of the right M1(arrowheads), indicative of cerebral vasculitis.Download figureDownload PowerPointFigure 4. A 39-year-old woman presented with hemiparesis of the right arm and leg, headache, and aphasia caused by multiple ischemic infarcts in the left middle cerebral artery territory as seen on diffusion-weighted magnetic resonance images suspected of having Sneddon syndrome. A, Postcontrast 3 T, 3-dimensional, T1-weighted volumetric isotropic turbo spin-echo acquisition vessel wall image (sequence adjusted from Qiao et al;20 repetition time/echo time, 1500/35 milliseconds) showing vessel wall enhancement of the right distal internal carotid artery (ICA). B through D, Corresponding 7 T, transverse, 3-dimensional magnetization preparation inversion recovery turbo spin-echo images before (B), directly after (C), and after 20 minutes (D) of contrast administration. B, Vessel wall lesions of the distal ICA on both sides. C, Enhancement of both the left and right distal ICA with even further enhancement after 20 minutes of contrast administration (D). The concentricity of the lesions and involvement of both ICAs can be indicative of moyamoya disease.Vulnerable PlaquesEnhancing and nonenhancing plaques are classifications based on overt imaging characteristics, but a characteristic that is probably clinically more relevant is whether the plaque can be described as vulnerable. Vulnerable plaques, that is, atherosclerotic “active” plaques at risk for rupture, are associated with a high risk of (recurrent) ischemic events.64 Similar to the characterization of carotid plaques, MRI might be used for the characterization of intracranial atherosclerotic plaques to assess the vulnerability of these plaques. For carotid plaques, plaque content differentiation includes the presence of calcifications, fibrin, lipid, vasa vasorum (plaque vascularization), and intraplaque hemorrhage.64 Detection of intraplaque hemorrhage in particular was found to be related to a higher recurrent stroke risk.65 In addition to plaque content, the presence of ulceration66 and the size of the atherosclerotic plaque67 were found to be correlated with prognosis. With further development of intracranial plaque imaging methods, we expect that characterization of the intracranial arterial plaque will also become possible in the next decade. Detection of intracranial plaque hemorrhage has already been shown to be possible intracranially38,51 with an MR direct thrombus imaging51 sequence for the middle cerebral artery (Table 1). In addition, several studies have investigated plaque vulnerability by means of contrast-enhanced sequences or used multiple contrast weightings (Figure 5).10,32,38,41,42,68 Similar to black-blood techniques, the most successful MRI sequences for the detection of vulnerable plaque in the carotid arteries, for instance, MR direct thrombus imaging51 combined with several image contrast weightings, should be optimized and tested for intracranial plaque characterization.Download figureDownload PowerPointFigure 5. A 31-year-old man with moyamoya disease. A 7 T, 3-dimensional magnetization preparation inversion recovery turbo spin-echo (A and D) and whole-brain T1-weighted (B and E) and T2-weighted (C and F) magnetization preparation inversion recovery turbo spin-echo images. The vessel wall of the left internal carotid artery (ICA; (white arrow on the right in A–C) is slightly more hyperintense compared with the right ICA (white arrow on the left in A–C) and basilar artery (black arrow in A–C). Both distal ICAs are relatively narrow compared with the basilar artery. The left middle cerebral artery is seen to be almost occluded, with thickened vessel wall (dashed arrow in D–F). The normal P1 segment of the posterior cerebral artery can also be assessed (white arrow in D–F). Reproduced from van der Kolk et al68 with kind permission of the publisher. Copyright © 2013 Springer Science+Business Media.Clinical ImplementationIn current clinical practice, it can be difficult to establish the cause of ischemic stroke in an individual patient, and differentiation between stroke subtypes is still an important issue. A method used to define the subtype of ischemic stroke is the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification system.1 Stroke subtypes defined by this system are strokes with a large-artery atherosclerotic cause, cardioembolic strokes, small-artery strokes, strokes with a determined cause, and strokes with an undetermined cause. This classification system has limitations, with many