low-grade carotid stenosis with intraplaque hemorrhage and expansive arterial remodeling is associated with a high relapse rate refractory to medical treatment.

Background: Carotid plaque characteristics considerably influence future risk of stroke. 2 However, the severity of stenosis does not accurately reflect plaque burden in patients 3 with expansive arterial remodeling. Objective: The present study aimed to determine the therapeutic outcome of 5 symptomatic carotid low-grade stenosis with vulnerable plaque based on MRI 6 characterization. 7 Methods: We studied 25 (male, n = 23; age, 74.2 ± 5.6 years) of 29 consecutive patients 8 with symptomatic carotid low-grade stenosis (< 50%) and both high-signal plaque and 9 expansive remodeling on T1-weighted MR images. The remaining four were excluded 10 due to impending stroke. A single antithrombotic and statin were administered and 11 recurrent ischemic stroke was treated with dual antithrombotics. We considered carotid 12 endarterectomy (CEA) when recurrence was refractory to aggressive medical treatment. 13 Results: During a 31.3 ± 16.4 month follow-up, 11 of the 25 patients developed a total 14 of 30 recurrent ischemic events (46.0% per patient-year). The patients’ characteristics 15 did not significantly differ between the groups with and without recurrence (n = 11 and 16 n = 14, respectively). Seven of 11 patients in the recurrence group who were treated 17 with CEA remained free of ischemic events during a postoperative follow-up of 19.1 ± 18 14.6 months. 19 Conclusion: Symptomatic low-grade carotid stenosis with vulnerable plaque confirmed 20 by MRI was associated with a high rate of stroke recurrence that was refractory to 21 aggressive medical treatment. However, CEA was safe and effective for such patients. Plaque characterization by MRI has potential for more accurate stroke risk stratification 23 in the management of carotid low-grade stenosis.


Introduction 1
Several large multicenter trials have established that carotid endarterectomy (CEA) 2 significantly reduces stroke risk compared with medical treatment in patients with 3 moderate to high-grade stenosis, but has no apparent benefit for those with low-grade 4 stenosis. 1-4 The assessment of stroke risk from carotid atherosclerosis and determination 5 of the need for surgical intervention in these randomized trials have been based on 6 angiographic or ultrasonographic evaluations of stenosis rates. 5 7 Accumulating evidence from recent vascular biology studies indicates that plaque 8 stability depending on several factors such as plaque components, volume and 9 morphological characteristics has a large impact on the risk of ischemic events from 10 atherosclerotic disease. 6-9 Therefore, the importance of evaluating not only the severity 11 of stenosis but also the vessel wall itself has been widely recognized for the 12 management of carotid stenosis. While carotid plaque is generally assessed by 13 ultrasonography, MRI has been gaining importance in the evaluation of carotid 14 atherosclerosis with recent technological developments in MRI equipment and imaging 15 sequencing. Many investigations have demonstrated that carotid MRI can accurately, 16 non-invasively and objectively provide a considerable amount of information about 17 carotid atherosclerosis. 10-13 Moreover, several reports have revealed an important 18 characteristic of plaque associated with a high risk of stroke on MR images. 14-17 19 Luminography does not necessarily reflect a significant burden of atherosclerotic 20 plaque due to expansive arterial remodeling in some patients with carotid stenosis. 18,19 21 Some might be at high risk for stroke even if angiography reveals only mild stenosis. 22 Therefore, meticulous risk assessment for ischemic events based on plaque evaluation 23 in addition to stenosis rates is significant, particularly for patients with low-grade 24 A Self-archived copy in Kyoto University Research Information Repository https://repository.kulib.kyoto-u.ac.jp angiography (DSA), computed tomography angiography (CTA), or magnetic resonance 1 angiography (MRA) using the North American Symptomatic Carotid Endarterectomy 2 Trial (NASCET) criteria. 3 All patients who were scheduled for CEA were also examined 3 by DSA. 4 Atherosclerotic plaque associated with intraplaque hemorrhage (IPH) was diagnosed 5 when axial T1-weighted 3D gradient-echo black blood (BB)-MRI images revealed high 6 signals relative to the sternocleidomastoid muscle. 22 Using axial MR images of areas 7 with the highest rate of stenosis, relative overall plaque MR signal intensity (roSI) was 8 measured and compared with plaque MRI signals as described. 23 9 Plaque in the patients who underwent CEA was characterized in more detail by 10 obtaining several additional MR images as follows. Plaque components were precisely 11 evaluated using 2D spin-echo BB-MRI, 23 plaque distribution was evaluated using 12 long-axis 3D inversion-recovery turbo field echo BB-MRI, 24  Medical treatment was started essentially with a single antithrombotic drug selected 1 from among aspirin, ticlopidine, clopidogrel and cilostazol. Concomitant hypertension, 2 hyperlipidemia or diabetes was also treated if necessary. Angiotensin II receptor 3 blockers and statins were preferentially administered to control hypertension and 4 hyperlipidemia, respectively. Dual antithrombotic therapy was administered if patients 5 developed recurrent ischemic stroke or had other ischemic vascular diseases. Ischemic 6 symptom recurrence was defined as TIA, MRI-confirmed infarct or amaurosis fugax in 7 the territory of the ipsilateral carotid artery. 8 We examined the implementation of CEA when patients presented with more than 9 two recurrences despite maximal medical therapy, had no other general problems and 10 consented to surgical intervention. Carotid endarterectomy proceeded under general 11 anesthesia and with intra-operative monitoring of somatosensory evoked potentials 12 (SEP) and near-infrared spectroscopy (NIRS). An internal shunt was selectively applied 13 when the SEP amplitude fell by > 50% or the value of NIRS dropped by > 20%. 14 Patients continued medical treatment with a single antithrombotic after undergoing 15

Statistical analysis 17
We investigated the number of recurrences during follow-up under medical treatment 18 (period between starting medical treatment and surgical intervention for patients who 19 underwent CEA). Differences in risk factors, medical regimen, roSI on BB-MRI, and 20 duration of follow-up were analyzed between groups with and without recurrence. We 21 also analyzed excised specimens, perioperative complications and postoperative 22 recurrence after CEA. 23 Numerical data are expressed as means ± standard deviation. Statistical significance 24 was calculated using Student's t test or Fisher's exact test and probability values of < 1 0.05 were considered significant. All data were statistically analyzed using SPSS 12.0 2 for Windows. patient-year. None of age, gender, hypertension, hyperlipidemia, diabetes, ischemic 9 heart disease, atrial fibrillation, medical regimen, roSI and duration of follow-up 10 significantly differed between the groups with (n = 11) and without (n = 14) recurrence 11 (Table 1). 12 Seven of 11 patients with recurrence underwent CEA 12.1 ± 10.4 days after the last 13 ischemic attack. The duration between the initial event and CEA was 35.4 ± 16.9 14 months and the number of recurrences before CEA was 3.29 ± 2.06. The findings of 15 DSA revealed a stenosis rate of 27.1 ± 9.9% and obvious ulceration in two patients. 16 Distinctly high signals of plaque were found on BB-MRI of all patients and the roSI 17 No ischemic events, including those in the contralateral hemisphere, were confirmed 2 during the postoperative follow-up period of 19.1 ± 14.6 months ( Table 2).  Neither atherosclerotic plaque along the ascending aorta on aortic MRI nor intracardiac 13 thrombus on transesophageal echocardiography was detected. Hyperlipidemia was 14 confirmed by laboratory data upon admission. 15 These findings indicated a diagnosis of recurrent ischemic events due to 16 artery-to-artery embolism originating from the carotid low-grade stenosis with unstable 17 plaque. The patient was administered with dual antithrombotics and statin. He was 18 readmitted to our hospital during July 2008 with right arm and leg weakness. Diffusion 19 weighted images demonstrated fresh multiple minor infarcts in the left hemisphere, 20 whereas left carotid angiography identified no appreciable elevation in the rate of 21

stenosis. 22
Because not even maximal medical therapy could prevent recurrent ischemic events, 23 the patient provided written informed consent to undergo CEA on the seventh day from 24 the last event. The postoperative course was uneventful and no new infarct was 1 demonstrated on diffusion weighted images obtained two days postoperatively. 2 Macroscopic observation of an excised specimen identified a disrupted fibrous cap (Fig.  3 1D) and histological assessment also revealed a ruptured fibrous cap and massive IPH. 4 The patient was treated only with aspirin after CEA and he has remained stroke-free at 5 14 months of follow-up. Carotid atherosclerosis has been widely evaluated by ultrasonography, as it is 9 noninvasive and can provide information about luminal stenosis and, to some extent, 10 plaque characteristics. 26, 27 However, limitations are associated with assessing the 11 carotid artery using this modality since discriminating IPH from a lipid core in 12 atherosclerotic plaque is less robust. Furthermore, plaque with dense calcification, 13 stenosis at higher locations and patients with a short neck cannot be fully imaged using 14 US and interpretation of the results is extremely operator-dependent. 28 be visualized as intense signals on T1-weighted MR images is generally accepted and 20 intense signals are generated more frequently by symptomatic, than asymptomatic 21 carotid plaque. 10, 15, 33 The MR signal of IPH is obviously high during the acute phase 22 and declines over time. 34 Prospective studies of carotid atherosclerosis have revealed 23 that stenosis with IPH detected by MRI confers higher risk for ischemic events than that 24 without IPH in either symptomatic 35 or asymptomatic patients. 17 Considering the many 1 reports describing MRI plaque evaluation of patients with carotid stenosis, regarding 2 plaques with high-signal on T1-weighted MR images as being unstable and at high risk 3 for ischemic stroke seems appropriate. 4 Few studies have examined the outcomes of medical treatment for symptomatic 5 low-grade carotid stenosis. The European Surgery Carotid Trial (ECST) reported only a 6 1.3% rate of ipsilateral ischemic strokes lasting beyond seven days in patients with 7 symptomatic mild (0-29%) carotid stenosis during three years of follow-up (0.43% per 8 year) when stroke recurrence assessment was based on merely stenosis rate evaluation. 36 9 Among patients with < 50% stenosis in the NASCET study, the 5-year rate of any 10 ipsilateral stroke was 18.7% in the medically treated group (3.74% per year). 3 Fritz et al. 11 reported an 8.6% rate of ischemic events including TIA and stroke during two years of 12 follow-up (4.3% per year) in 35 medically treated symptomatic patients with low grade 13 stenosis (< 50%) or ulcerated plaques. 37 Although the rate of recurrent ischemic events 14 has some variation among these studies, probably because of differences in the 15 definition of ischemic events or rate of stenosis, they nevertheless concluded that 16 patients with symptomatic low-grade stenosis did not benefit from CEA because the 17 outcome of medical treatment was favorable. 18 To the best of our knowledge, only Altaf et al. 20 prospectively analyzed recurrence 19 risk based on MRI plaque evaluation in addition to rates of stenosis among patients with 20 symptomatic carotid low-grade stenosis. They showed that the rates of recurrent 21 ischemic events in patients with symptomatic mild stenosis (30-49%) treated medically 22 during a median follow-up period of 28 months were 25% and 10% in plaques with IPH 23 detected or not by MRI, respectively. They concluded that carotid plaques with intense 24 signals on T1-weighted MR images are at higher risk for ischemic recurrence even 1 when the rate of stenosis is not significant. 2 The present retrospective analysis found ischemic recurrence in 44% of medically 3 treated patients who had high-signal plaque and expansive remodeling on T1-weighted 4 carotid MR images, and an annual recurrence rate reached 46.0% per patient-year. A 5 detailed comparative discussion is limited because reports on outcomes of medical 6 treatment for symptomatic mild stenosis based on MRI plaque evaluation are scarce. 7 Nonetheless, we can offer some explanations for the remarkably high rate of ischemic 8 recurrence in this study. Recurrence in our report included both stroke and TIA. The 9 recurrence rate excluding TIA in this study was still high at 36.8% per patient-year. 10 Stroke was diagnosed based on diffusion-weighted images regardless of the severity of 11 ischemic symptoms. The chances of an oversight in ischemic events seemed low 12 because patients were regularly followed up at a single institution. More importantly, 13 the fact that our study patients had not only carotid plaque with IPH detected by MRI 14 but also expansive remodeling might explain why our patients were at higher risk for 15 ischemic events than those described by Altaf et al. Expansive plaque remodeling in 16 patients with coronary artery atherosclerosis has been considered a major factor in 17 vulnerable plaque. 18, 19 A study using a murine model of carotid artery lesions found that 18 the macrophages characteristic of vulnerable plaques assist in expansive remodeling by 19 promoting matrix degeneration through matrix metalloproteinase secretion. 38 A recent 20 study of human carotid atherosclerosis using multidetector CT angiography also found 21 that the incidence of expansive remodeling is much higher among symptomatic, than 22 asymptomatic carotid plaques and that the extent of expansive remodeling correlated 23 with atherosclerotic plaque vulnerability. 21 24 With respect to the safety of CEA for treating symptomatic carotid stenosis, large 1 clinical trials have shown that the perioperative complication rate does not depend on 2 the rate of carotid stenosis. Rates of death, disabling stroke, or any other stroke 3 producing symptoms for more than seven days in the ECST report were 3.7% (70-99%) 4 and 2.3% (0-29%) for severe and mild stenosis, respectively, with no statistically 5 significant difference. 36 In addition, NASCET did not find a significant difference in the 6 rate of disabling stroke at 30 days between groups with moderate (2.8%) and severe 7 (3.0%) stenosis. 39 The fact that diffusion-weighted MR imaging in the present study 8 confirmed the absence of perioperative ischemic events demonstrated the safety of CEA 9 for patients with symptomatic low-grade stenosis. None of the patients described herein 10 has presented with TIA or any stroke, even though they were treated with only single 11 antiplatelet drugs after CEA for a median postoperative follow-up of 19 months. 12 Therefore, our results might also indicate the relevance of CEA for patients with 13 symptomatic low-grade stenosis accompanied by vulnerable plaque confirmed by MRI, 14 although this assertion is limited by the number of patients studied and the short period 15 of postoperative follow-up. 16 Carotid endarterectomy was adopted as a therapeutic option for patients who were 17 refractory to multidisciplinary medical treatment in this study. However, we could not 18 find distinctive differences in the patients' characteristics and MRI findings between the 19 recurrence and non-recurrence groups, which remains an issue. To discriminate patients 20 with low-grade carotid stenosis refractory to medical treatment before recurrences or 21 screen extremely vulnerable plaques with diagnostic imaging remains challenging. In 22 addition to IPH and expansive arterial remodeling, fibrous cap rupture is also an 23 important feature of vulnerable plaques. In fact, a prospective study of asymptomatic 24 carotid stenosis with MRI plaque evaluation has demonstrated that ruptured fibrous caps 1 more significantly influence ischemic events than IPH. 17 Precise assessment of fibrous 2 cap status would thus seem key to discriminating extremely vulnerable plaques. Several 3 studies of MRI plaque characterization have shown that ruptured or thin fibrous caps 4 can be detected as "dark bands" on TOF images 25, 32, 40 and have demonstrated a 5 moderate to good correlation between MRI findings and CEA specimens for diagnosing 6 ruptured caps. However, these reports also pointed out the limitation of fibrous cap 7 estimation by MRI due to poor image quality when the internal carotid artery is 8 obviously angulated or tortuous or when plaques are densely calcified. In fact, 18-34% 9 of the patients who underwent carotid MRI were excluded from analysis of the fibrous 10 cap in these studies. The present study found that the preoperative estimation of fibrous 11 cap from TOF MR images agreed with histological findings of excised plaques in only 4 12 of 7 patients. Thus, MRI is not yet satisfactory for fibrous cap assessment in the clinical 13 setting. Improved spatial-and contrast-to-noise ratios, better spatial resolution using a 14 higher magnetic field and faster imaging procedures will allow the identification of finer 15 structures within atherosclerotic arterial walls 41 and enable more precise diagnosis of 16 high-risk plaques for recurrent ischemic events. 17 The preventive effect against ischemic stroke for low-grade vulnerable stenosis 18 between CEA and medical treatment should be compared. While anti-thrombotic drugs 19 have been widely used to prevent strokes, 42 more aggressive medical treatment with 20 dual anti-platelet medication with aspirin and clopidogrel is apparently more effective 21 than aspirin alone. 43 In addition, statins seem to be appreciably effective against many 22 disorders caused by systemic atherosclerosis. The ability of aggressive lipid-lowering 23 therapy with statins to reduce the incidence of ischemic stroke has been verified by 24 several large randomized studies. 44,45 In addition, high-resolution MRI studies have 1 demonstrated that statins suppress, stabilize and reduce carotid atherosclerotic 2 plaque. 46-48 Thus, the outcome of medical treatment for carotid stenosis should be 3 significantly improved by early multifaceted medical intervention for patients with 4 high-risk atherosclerotic disease. Comparing the long-term results between CEA and 5 aggressive medical treatment is mandatory to validate the use of CEA for low-grade 6 high-risk stenosis. 7 One limitation of the present study is that it was a retrospective analysis of a small 8 patient cohort over a relatively short follow-up period. Importantly, our patients did not 9 have symptomatic low-grade carotid stenosis but had plaque with positive remodeling 10 that was evident on high-signal on T1-weighted images. Therefore, the present findings 11 are not applicable to symptomatic low-grade stenosis in general. Nevertheless, our 12 findings promote awareness of the limitations of future stroke risk assessment based 13 solely on the severity of carotid stenosis. 14

Conclusion 15
Among patients with symptomatic low-grade carotid stenosis, those with plaque 16 containing IPH demonstrated by MRI undergoing outward remodeling were associated 17 with a high rate of stroke recurrence regardless of medical treatment. Carotid 18 endarterectomy prevented further stroke in patients who were refractory to aggressive 19 medical therapy and was validated for treating a subset of patients with low-grade, but 20 high-risk carotid stenosis determined by plaque evaluation.   Table 1. Characteristics of the patients with symptomatic low-grade carotid stenosis.
roSI, relative overall signal intensity on axial T1-weighted images of whole carotid plaque except for lumen at maximal stenosis relative to adjacent sternocleidomastoid muscle.; 23 f/u, follow-up.