Incidence of New Brain Lesions After Carotid Stenting With and Without Cerebral Protection
Background and Purpose— Diffusion-weighted imaging (DWI) may be a useful tool to evaluate the efficacy of cerebral protection devices in preventing thromboembolic complications during carotid angioplasty and stenting (CAS). The goals of this study were (1) to compare the frequency, number, and size of new DWI lesions after unprotected and protected CAS; and (2) to determine the clinical significance of these lesions.
Methods— DWI was performed immediately before and within 48 hours after unprotected or protected CAS. Clinical outcome measures were stroke and death within 30 days.
Results— The proportion of patients with any new ipsilateral DWI lesion (49% versus 67%; P<0.05) as well as the number of new ipsilateral DWI lesions (median=0; interquartile range [IQR]=0 to 3 versus median=1; IQR=0 to 4; P<0.05) were significantly lower after protected (n=139) than unprotected (n=67) CAS. The great majority of these lesions were asymptomatic and less than 10 mm in diameter. Although there were no significant differences in clinical outcome between patients treated and not treated with protection devices (7.5% versus 4.3%, not significant), the number of new DWI lesions was significantly higher in patients who developed a stroke (median=7.5; IQR=1.5 to 17) than in patients who did not (median=0; IQR=1 to 3.25; P<0.01).
Conclusions— The use of cerebral protection devices significantly reduces the incidence of new DWI lesions after CAS of which the majority are asymptomatic and less than 10 mm in diameter. The frequent occurrence of these lesions and their close correlation with the clinical outcome indicates that DWI could become a sensitive surrogate end point in future randomized trials of unprotected versus protected CAS.
Carotid endarterectomy is currently widely performed for severe carotid artery stenosis. However, the benefit of carotid endarterectomy is highly dependent on a low risk of procedural neurological complications and is eliminated when the combined 30-day stroke and death rates exceed approximately 5% to 7% for symptomatic and 3% or even lower for asymptomatic patients, respectively.1,2 Because higher morbidity and mortality rates have been reported when carotid endarterectomy is used in everyday clinical practice,3 carotid angioplasty and stenting (CAS) might become an attractive alternative treatment strategy. Although evidence is accumulating that CAS can be performed with acceptable complication rates,4,5 fear of distal embolization of plaque fragments to the brain has generated great concern regarding the safety of this technique. Therefore, recent efforts have focused on the development of cerebral protection devices aimed at preventing the passage of embolic material into the cerebral vasculature. Although the concept of cerebral protection during CAS is appealing and has indirectly been supported by several case series and stent registries,4,5 no randomized study has yet been conducted to investigate the clinical efficacy of this approach. Moreover, the use of cerebral protection devices increases the intervention time, the complexity and cost of the procedure, and may also have a negative influence on embolization rates themselves.6,7
Because clinical events after CAS are relatively uncommon, additional surrogate markers for clinical stroke that occur at a greater frequency would have considerable use to evaluate the efficacy of protection devices. Against the background of a high incidence of clinically silent emboli occurring during CAS detected by diffusion-weighted imaging (DWI),8–14 this imaging modality could become a useful tool in this scenario. In fact, DWI is currently the most sensitive tool to detect early cerebral ischemia15 and offers the possibility of making even small and thus asymptomatic lesions visible shortly after their emergence.16 Although several DWI studies have documented a high incidence of silent ischemia after either unprotected or protected CAS,8–14 until now, only one small study published recently has directly compared the frequency of new DWI lesions between protected and unprotected CAS.17 Moreover, the clinical significance of these lesions has not yet been established in a larger patient population. Therefore, the goals of the present study were (1) to compare the frequency, number, and size of new DWI lesions after unprotected and protected CAS; and (2) to determine the clinical significance of these lesions.
From April 1999 to October 2005, consecutive patients with high-grade carotid stenosis (≥70% in symptomatic patients and ≥90% in asymptomatic patients assessed with ultrasound) were treated with CAS after a prospective protocol at our institution. The severity of carotid stenosis was evaluated by measuring the peak systolic velocity with angle correction at the narrowest point of stenosis. A stenosis was classified ≥70% if the peak systolic velocity was greater than 210 cm/second and ≥90% if it was >300 cm/second, respectively. A carotid stenosis was considered symptomatic if the patient had experienced an ipsilateral ocular or cerebral (permanent or transient) ischemic event within the past 6 months. Those patients without contraindications for magnetic resonance imaging (MRI) received preinterventional and postinterventional DWI scans of the brain. All patients were informed of the investigative nature of CAS and gave their written consent. Our Institutional Ethics Review Board had approved our CAS protocol.
Carotid Stent Protocol
All patients were treated with carotid angioplasty with stenting according to a standardized protocol described in detail recently.18 Initially, all CAS procedures had been performed without cerebral protection devices. When cerebral protection devices became available, the choice of which type of device to use, if any, depended on the personal preference of the interventional neuroradiologist performing the procedure.
In all patients, MRI scans were obtained immediately before and within 48 hours after the intervention. MRI was performed by echoplanar imaging using a 1.5-T MRI system (Siemens Magnetom Vision or Sonata; Siemens). Multislice diffusion-weighted single-shot echoplanar images were acquired in all patients while using the following parameters: repetition time (TR)=0.8 ms; echo time (TE)=123 ms; acquisition time-4 seconds; and b=1100 s/mm2. Diffusion sensitivity was in the slice selection direction and hence perpendicular to the imaging plane. The number of measurements was five, the first run was omitted, and the remaining four were added to create an average image with improved signal-to-noise ratio. The conventional MRI sequences included T2-weighted fluid-attenuated inversion recovery turbo spin echo images (TR=9000 ms; inversion time=2200 ms; TE=119 ms). A magnetic resonance angiography was performed in all subjects before CAS using either a time-of-flight technique or with a heavily T1-weighted, contrast bolus-enhanced three-dimensional gradient echo sequence (TR=3.2 ms; TE=1.2 ms; flip angle=30°; field of view=300; 60 to 70-mm slice thickness; 36 partitions).
MR image analysis was performed jointly by a neuroradiologist (U.E.) and a neurologist (A.K.) who both had extensive experience with the interpretation of DWI scans. With the exception of those patients who had developed a stroke during CAS and had also been examined physically by A.K., both readers were blind to the clinical data. In addition, A.K. was blinded to the use of cerebral protection devices. Abnormal DWI lesions were identified in each patient by visual inspection of the MR scans. New DWI lesions were determined by slice-to-slice comparison of the DWI images between both scanning sessions. In case of dissent, a second neuroradiologist reviewed the images and a decision was made by consent among all three reviewers. Enlargement of a previous DWI lesion was not considered as a new ischemic lesion. All new DWI lesions were described by their number, location in the brain, and their maximal diameter (given in millimeters and classified as <10 mm, 10 to 20 mm, or >20 mm). Large confluent lesions and territorial infarctions were noted separately. The preinterventional magnetic resonance angiographies were used to decide if the new DWI lesions were inside or outside the vascular territory of the treated artery. This was done by (1) determining the distribution of the lesions within the different vascular territories (anterior or posterior circulation; ipsilateral or contralateral to the treated artery) and (2) visualizing collateral blood flow patterns within the circle of Willis.
With respect to the number, size, and location of the DWI lesions, interobserver agreements were assessed with kappa statistics. A κ of 1 indicates perfect agreement, whereas zero shows only chance agreement, excellent agreement refers to values greater than 0.80, whereas 0.61 to 0.80 indicates substantial agreement, and 0.41 to 0.60 indicates moderate agreement.
Data Collection and Clinical Evaluation
History-taking and neurological examination were performed for each patient by one of three stroke neurologists (A.K., K.G., or J.B.S.) before CAS and additional neurological examinations were performed by one of two board-certified neurologists (A.K. and J.B.S.) the day after CAS and at day 30.
The following cerebrovascular risk factors were recorded using history or direct measurements: hypertension (blood pressure ≥160/90 mm Hg measured on repeated occasions), diabetes mellitus (HbA1c >6.5% or fasting blood glucose >120 mg/dL), hyperlipidemia (fasting serum cholesterol levels >220 mg/dL), smoking (current or within the previous year), previous transient ischemic attacks and strokes, coronary artery disease (angina, myocardial infarction, percutaneous transluminal angioplasty, or surgery), and the presence of contralateral carotid disease (assessed with ultrasound).
Definitions of Clinical Outcome Measures
The clinical outcome measures were minor/major stroke or death within 30 days19 and were defined as follows:
Any new neurological deficit (either ocular or cerebral) that persisted for more than 24 hours and that either resolved completely within 30 days or increased the National Institutes of Health Stroke Scale <3 points.
Any new neurological deficit that persisted after 30 days or increased the National Institutes of Health Stroke Scale by >3 points.
Continuous values were expressed as mean±SD and nominal variables as count and percentages. Median values and the interquartile range were computed as appropriate. For comparisons of categorical data, 2-tailed χ2 statistics with Yates correction and univariate Fisher exact test were used. The Fisher exact test was used when the predicted contingency table cell values were <5. Analyses of continuous variables between the cohorts were performed with an unpaired Student t test. Because the imaging data were not distributed normally, differences between both groups were tested by using the Mann-Whitney U statistic. A value of P<0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed with SPSS (version 12; SPSS Inc).
From April 1999 to October 2005, a total of 353 consecutive patients had been treated with CAS at our institution. Within this series, 206 patients (152 male and 54 female; mean age=69±9 years; range=42 to 89 years) comprised the study population for this analysis. The remaining 147 patients were excluded because they had declined or were unable (eg, as a result of claustrophobia) to participate in this substudy (56%) or had had contraindications for MRI examinations (44%).
The demographic and clinical characteristics of patients treated with and without cerebral protection devices were similar and are summarized in Table 1. In both groups, the demographic and clinical characteristics were also comparable between those patients treated before and those treated after protection devices had become available.
MRI Lesion Load and Protection Devices
Twenty-three patients were treated without cerebral protection devices before these had become available, 44 patients were treated without cerebral protection devices after these had become available, and in 139 patients, filter-type embolic protection devices were used during the CAS procedures. According to physician preference and commercial availability, four different cerebral protection devices were used in this study: Neuroshield, n=31 (MedNova); Angioguard, n=11 (Cordis J&J); Emboshield, n=61 (Abbott); and Filterwire, n=36 (Boston Scientific).
There was overall agreement between both reviewers as to the number (κ=0.94 for interobserver agreement, 95% CI=0.91 to 0.97), size (κ=0.93 for interobserver agreement, 95% CI=0.87 to 0.99), and location (κ=0.82 for interobserver agreement, 95% CI=0.72 to 0.91) of new DWI lesions. Before the procedure, DWI revealed ipsilateral cerebral lesions in 15 (22%) of the patients treated without and in 22 (16%) of the patients treated with cerebral protection devices (not significant). The proportion of patients with any new ipsilateral DWI lesion in the group of unprotected CAS was 61% in those patients treated before protection devices had become available and 70% after these had become available (not significant). Moreover, the number of new ipsilateral DWI lesions was comparable between these two groups (median=1; interquartile range [IQR]=0 to 6 versus median=1.5; IQR=0 to 4; not significant) so that they were combined for further analysis.
The proportion of patients with any new ipsilateral DWI lesion(s) after protected CAS was significantly lower than after unprotected CAS (49% versus 67%; P<0.05). In addition, the total number of new ipsilateral DWI lesions was significantly lower after protected CAS than after unprotected CAS (median=0; IQR=0 to 3; versus median=1; IQR=0 to 4; P<0.05), and in both groups, the vast majority of new DWI lesions had a diameter of <10 mm (Table 2).
No statistical correlations were found between the incidence or number of any new ipsilateral DWI lesion and the type of protection device used, the degree of stenosis, or the presence of a contralateral stenosis. A total of 14 (21%) patients treated without and 24 (17%) of the patients treated with cerebral protection devices developed new DWI lesions outside the vascular territory of the target lesion (P=0.1).
MRI Lesion Load and Stroke Outcomes
The neurological complications within 30 days for patients with and without cerebral protection are summarized in Table 3. All minor or major strokes occurred in the territory of the treated artery and within 24 hours after CAS. There were no significant differences in the overall clinical complication rates between patients treated without versus those treated with protection (7.5% versus 4.3%, not significant). However, in patients who had developed a minor or major stroke, the number of new DWI lesions (median=7.5; IQR=1.5 to 17) was significantly higher than the number of new DWI lesions in patients without neurological deficits (median=1; IQR=1 to 3.0; P<0.01). Both patients with major stroke had multiple DWI lesions larger than 20 mm in diameter, whereas all patients with minor stroke had had multiple (range=2 to 15) new DWI lesions smaller than 10 mm in diameter.
Our study is the largest to date directly comparing the number, size, and location of new DWI lesions with respect to unprotected and protected CAS. We demonstrated that the application of a filter systems during CAS significantly reduced the proportion of patients with any new ipsilateral DWI lesion. Moreover, patients treated with cerebral protection had significantly fewer new DWI lesions than unprotected patients, conferring with results of a recent small study.17
Also consistent with investigation by others was our finding that the majority of new DWI lesions were smaller than 10 mm and were asymptomatic.8–14 It must be kept in mind that these lesions do not necessarily represent irreversible brain damage.10 On the other hand, in our study, patients who developed a stroke had significantly more new DWI lesions than those who did not, indicating a strong association between DWI lesions and ischemic injury. To the best of our knowledge, a similar finding has not been reported before.
To date, no randomized trial has shown that cerebral protection devices reduce the incidence of stroke associated with CAS. Using our data at least 1500 patients would have to be randomized to detect a statistically significant difference in stroke and death rates between patients treated with unprotected versus protected CAS.
However, because DWI lesions are frequent and more common in those who develop stroke after CAS, it appears that DWI could be a sensitive surrogate end point for ischemic injury if the risk of stroke associated with these lesions could be quantified. Using our data and the proportion of patients with any new DWI lesion as the outcome measure, only 120 patients would have to be randomized to detect a statistically significant difference between patients treated or not treated with filter devices. Such a randomized study would be useful to overcome the limitations of this observational study, which was subject to some selection bias and operator improvements through experience and was not completely blind with respect to the assessment of the DWI lesions.
Finally, the high incidence of new brain lesions within the vascular territory of the treated carotid artery found in the group of protected patients with CAS documents that dislodgement of a large number of embolic particles to the brain is not prevented by the use of currently available filter devices. It is possible that (1) many emboli are too small to be captured by the filter devices; (2) various protection devices do not completely cover the internal carotid artery allowing emboli to pass by; or (3) the manipulation of catheters and guide wires within the internal carotid artery might have produced DWI lesions before deployment of the protection devices.20 Aside from its use to quantitatively evaluate ischemic injury after CAS, DWI could also be used to improve and develop new protection devices.
New cerebral lesions detected with diffusion-weighted MRI are frequent after both unprotected and protected carotid angioplasty and stenting, the majority of which are clinically silent. The use of cerebral protection devices significantly reduces the number of these lesions. The frequent occurrence of these lesions as well as their close correlation with the clinical outcome stresses the potential use of DWI to quantitatively evaluate ischemic injury after carotid angioplasty and stenting.
- Received December 19, 2005.
- Revision received April 9, 2006.
- Accepted May 23, 2006.
Biller J, Feinberg WM, Castaldo JE, Whittemore AD, Harbaugh RE, Dempsey RJ, Caplan LR, Kresowik TF, Matchar DB, Toole J, Easton JD, Adams HP Jr, Brass LM, Hobson RW 2nd, Brott TG, Sternau L. Guidelines for carotid endarterectomy: a statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke. 1998; 29: 554–562.
Abbott AL, Chambers BR, Stork JL, Levi CR, Bladin CF, Donnan GA. Embolic signals and prediction of ipsilateral stroke or transient ischemic attack in asymptomatic carotid stenosis: a multicenter prospective cohort study. Stroke. 2005; 36: 1128–1133.
Chaturvedi S, Aggarwal R, Murugappan A. Results of carotid endarterectomy with prospective neurologist follow-up. Neurology. 2000; 55: 769–772.
Kastrup A, Groschel K, Krapf H, Brehm BR, Dichgans J, Schulz JB. Early outcome of carotid angioplasty and stenting with and without cerebral protection devices: a systematic review of the literature. Stroke. 2003; 34: 813–819.
Vos JA, Van Den Berg JC, Ernst SM, Suttorp MJ, Overtoom TT, Mauser HW, Vogels OJ, van Heesewijk HP, Moll FL, van der Graaf Y, Mali WP, Ackerstaff RG. Carotid angioplasty and stent placement: comparison of transcranial Doppler US data and clinical outcome with and without filtering cerebral protection devices in 509 patients. Radiology. 2005; 234: 493–499.
Hauth EA, Jansen C, Drescher R, Schwartz M, Forsting M, Jaeger HJ, Mathias KD. MR and clinical follow-up of diffusion-weighted cerebral lesions after carotid artery stenting. AJNR Am J Neuroradiol. 2005; 26: 2336–2341.
Jaeger HJ, Mathias KD, Drescher R, Hauth E, Bockisch G, Demirel E, Gissler HM. Diffusion-weighted MR imaging after angioplasty or angioplasty plus stenting of arteries supplying the brain. AJNR Am J Neuroradiol. 2001; 22: 1251–1259.
Poppert H, Wolf O, Resch M, Theiss W, Schmidt-Thieme T, Graefin von Einsiedel H, Heider P, Martinoff S, Sander D. Differences in number, size and location of intracranial microembolic lesions after surgical versus endovascular treatment without protection device of carotid artery stenosis. J Neurol. 2004; 251: 1198–1203.
Roh HG, Byun HS, Ryoo JW, Na DG, Moon WJ, Lee BB, Kim DI. Prospective analysis of cerebral infarction after carotid endarterectomy and carotid artery stent placement by using diffusion-weighted imaging. AJNR Am J Neuroradiol. 2005; 26: 376–384.
van Everdingen KJ, van der Grond J, Kappelle LJ, Ramos LM, Mali WP. Diffusion-weighted magnetic resonance imaging in acute stroke. Stroke. 1998; 29: 1783–1790.
Cosottini M, Michelassi MC, Puglioli M, Lazzarotti G, Orlandi G, Marconi F, Parenti G, Bartolozzi C. Silent cerebral ischemia detected with diffusion-weighted imaging in patients treated with protected and unprotected carotid artery stenting. Stroke. 2005; 36: 2389–2393.
Mathur A, Roubin GS, Iyer SS, Piamsonboon C, Liu MW, Gomez CR, Yadav JS, Chastain HD, Fox LM, Dean LS, Vitek JJ. Predictors of stroke complicating carotid artery stenting. Circulation. 1998; 97: 1239–1245.