Author + information
- Received September 19, 2011
- Revision received November 23, 2011
- Accepted November 29, 2011
- Published online April 10, 2012.
- Klaudija Bijuklic, MD,
- Andreas Wandler, MD,
- Fadia Hazizi, MD and
- Joachim Schofer, MD, PhD⁎ ()
- ↵⁎Reprint requests and correspondence:
Prof. Dr. med. Joachim Schofer, Medical Care Center Prof. Mathey, Prof. Schofer, Hamburg University Cardiovascular Center, Wördemannsweg 25-27, 22527 Hamburg, Germany
Objectives The objective of this study was to compare the cerebral embolic load of filter-protected versus proximal balloon–protected carotid artery stenting (CAS).
Background Randomized trials comparing filter-protected CAS with carotid endarterectomy revealed a higher periprocedural stroke rate after CAS. Proximal balloon occlusion may be more effective in preventing cerebral embolization during CAS than filters.
Methods Patients undergoing CAS with cerebral embolic protection for internal carotid artery stenosis were randomly assigned to proximal balloon occlusion or filter protection. The primary endpoint was the incidence of new cerebral ischemic lesions assessed by diffusion-weighted magnetic resonance imaging. Secondary endpoints were the number and volume of new ischemic lesions and major adverse cardiovascular and cerebral events (MACCE).
Results Sixty-two consecutive patients (mean age: 71.7 years, 76.4% male) were randomized. Compared with filter protection (n = 31), proximal balloon occlusion (n = 31) resulted in a significant reduction in the incidence of new cerebral ischemic lesions (45.2% vs. 87.1%, p = 0.001). The number (median [range]: 2 [0 to 13] vs. 0 [0 to 4], p = 0.0001) and the volume (0.47 [0 to 2.4] cm3 vs. 0 [0 to 0.84] cm3, p = 0.0001) of new cerebral ischemic lesions were significantly reduced by proximal balloon occlusion. Lesions in the contralateral hemisphere were found in 29.0% and 6.5% of patients (filter vs. balloon occlusion, respectively, p = 0.047). The 30-day MACCE rate was 3.2% and 0% for filter versus balloon occlusion, respectively (p = NS).
Conclusions In this randomized trial of patients undergoing CAS, proximal balloon occlusion as compared with filter protection significantly reduced the embolic load to the brain.
Stroke is the third leading cause of death in industrialized countries and the major cause of functional impairment (1). Carotid artery stenosis is an important cause of stroke (2). Carotid endarterectomy (CEA) has been established as an effective treatment for symptomatic as well as for asymptomatic patients (3). Carotid artery stenting (CAS) has emerged as an alternative to CEA (4). Current data support the use of embolic protection devices for CAS (5). Randomized trials comparing CAS with CEA revealed lower periprocedural stroke rates with surgery (6–10). As protection devices for CAS, filters were used in these trials.
Proximal balloon occlusion is an alternative to filter protection, which, by occluding the external and common carotid artery (CCA), induces reversed flow in the target vessel before the lesion is crossed and stented. Whether this results in a more effective cerebral embolic protection than a filter device has never to our knowledge been studied in a randomized trial. Diffusion-weighted magnetic resonance imaging (DW-MRI) has been shown to be a sensitive tool in identifying new ischemic cerebral lesions caused by emboli during CAS (11–13) and can therefore serve as a surrogate endpoint. The primary objective of the present study was to compare the incidence, the number, and volume of new cerebral ischemic lesions after filter-protected versus proximal balloon–protected CAS.
The PROFI (Prevention of Cerebral Embolization by Proximal Balloon Occlusion Compared to Filter Protection During Carotid Artery Stenting) study is a prospective randomized, single-center trial. The study was approved by the Freiburg ethics committee (clinic trial code 010/1707). At that time, the operator had experience in CAS in more than 1,000 patients. All patients gave their written informed consent. Between May 2010 and August 2011, eligible patients were randomized in a 1:1 fashion to get either filter or proximal balloon protection for the CAS procedure. Before and 12 to 24 h after CAS, a cerebral DW-MRI was performed.
The primary endpoint was the incidence of new cerebral ischemic lesions as assessed by DW-MRI. Secondary endpoints were the number and volume of new cerebral ischemic lesions, major adverse cardiovascular and cerebral events (MACCE) (defined as death, stroke, myocardial infarction, and vascular complications) at 30 days, bleeding complications, and device crossover.
The inclusion and exclusion criteria are summarized in Table 1.
Patients were on dual antiplatelet therapy (aspirin 100 mg and clopidogrel 75 mg/day) before the intervention and for 4 weeks after. Carotid artery stenting was performed by a single experienced operator (J.S.) (14) as described in detail previously (11). The protocol specified use of the Cristallo ideale stent (Invatec, Roncadelle, Italy), the Emboshield Protection System for filter protection (Abbott Vascular, Abbott Park, Illinois), and the MO.MA ultra system (Invatec) for proximal balloon occlusion, respectively.
Diffusion-weighted magnetic resonance imaging
Cerebral DW-MRI scans were obtained before and 12 to 24 h after CAS using a 1.5-T scanner (Magnetom Sonata, Siemens, Erlangen, Germany). Images were reviewed by 2 independent physicians (A.W. and F.H.), both blinded to the protection device used. Echo planar imaging with the following parameters was used: TR = 3,000 ms, TE = 84 ms, 19 slices with a slice thickness of 6 mm, field of view = 230 mm, diffusion values b = 0, 500, 1,000 s/mm2, fat saturation, time of acquisition = 71 s. Additionally, apparent diffusion coefficient maps were obtained. A new lesion after CAS was defined as a focal hyperintense area detected by the fluid-attenuated inversion recovery sequence, corresponding to a restricted diffusion signal in the diffusion-weighted imaging sequence, confirmed by apparent diffusion coefficient mapping to rule out a shine-through artifact. The number, volume, and location of new ischemic lesions on DW-MRI at follow-up were assessed.
Patients underwent neurological examination by an independent neurologist prior to carotid artery stenting and before discharge and were monitored for at least 24 h. A follow-up visit with neurological evaluation and a duplex ultrasound of the target vessel was scheduled at 30 days.
Patients were considered symptomatic if they had an ipsilateral neurological ischemic event within 6 months before the procedure.
Diameter stenosis for patient screening was determined by ultrasound using the peak systolic velocity ratio, the ratio of the peak systolic velocity in the internal carotid artery to the peak systolic velocity in the distal CCA, with a value of <2.0 for <50% stenosis, >4.0 for 70% stenosis with the additional value of >5.0 for 90% stenosis but less than near occlusion. Diameter stenosis for evaluation of the procedural result was determined angiographically by visual estimate before and after stenting.
Procedural success was defined as residual stenosis after stenting ≤30% and the absence of complications.
Stroke was defined as a new neurological deficit with focal symptoms and signs consistent with focal cerebral ischemia, lasting at least 24 h. Stroke was considered minor if a neurological deficit resolved completely within 30 days or did not lead to an impairment in daily activities as judged by the independent neurologist. Otherwise, stroke was defined as major.
Balloon intolerance was defined as transient neurological symptoms during balloon occlusion which promptly disappeared after balloon deflation.
Myocardial infarction was defined as an increase in creatine kinase-MB 3 times the upper reference limit combined with electrocardiographic evidence of ischemia or symptoms consistent with myocardial ischemia.
Bleeding complications were defined according to the Thrombolysis In Myocardial Infarction criteria (15).
Cerebral ischemic lesions were defined as ipsilateral, if present in the hemisphere supplied by the treated carotid artery, otherwise as contralateral.
The type of aortic arch was defined as:
• Type I, if the vertical distance from the origin of the innominate artery to the top of the arch is <1 diameter of the left CCA.
• Type II, if the vertical distance from the origin of the innominate artery to the top of the arch is between 1 and 2 left CCA diameters.
• Type III, if the vertical distance from the origin of the innominate artery to the top of the arch is >2 left CCA diameters.
The level of stroke severity was measured by the National Institutes of Health Stroke Scale scoring system:
• 0 = no stroke
• 1 to 4 = minor stroke
• 5 to 15 = moderate stroke
• 16 to 20 = moderate/severe stroke
• 21 to 42 = severe stroke
Analyses were performed on the intention-to-treat basis. Continuous variables were presented as mean ± SD, and Mann-Whitney and Student t tests were used, as appropriate. Categorical variables were compared using Fisher exact test. For statistical analysis, Prism 3.0 (GraphPad Software, La Jolla, California) was used. Values of p < 0.05 were considered to indicate statistical significance.
Characteristics of patients
Of 84 consecutive patients, 62 fulfilled the study criteria and were randomized to either filter protection (n = 31) or proximal balloon occlusion (n = 31) (Fig. 1).
Demographic, clinical, and lesion characteristics of the patients, as summarized in Table 2, were not different between the 2 groups. The type of aortic arch did differ significantly between the groups (type I: 23% vs. 35%, type II: 52% vs. 25%, and type III: 9% vs. 40% in the filter vs. balloon occlusion group, respectively, p = 0.03).
Procedural duration was significantly longer in the balloon occlusion group compared with the filter group (30 ± 8 min vs. 22 ± 7 min, p = 0.003). Balloon intolerance was observed in 3 of 31 patients (6.5%); the procedure, however, could be concluded under cerebral protection in all cases. In none of the patients did we have to cross over to the filter or use a combination of a filter and a balloon.
The 22 patients who did not meet the inclusion criteria were entered in a registry and underwent the same protocol including the DW-MRI if not contraindicated. The reasons for excluding patients from randomization are summarized in Figure 1. Clinical and duplex ultrasound follow-up at 30 days was performed in all patients. There were no clinical events after hospital discharge, and the stent was patent in all patients.
The incidence of new cerebral ischemic lesions per patients in DW-MRI was significantly higher in the filter group compared with the balloon occlusion group (87.1% vs. 45.2%, p = 0.001) (Fig. 2). Results for symptomatic, asymptomatic, and age >80 years patients are shown in Figure 2. Lesions were located in the ipsilateral hemisphere in all patients with positive findings and, in addition, in the contralateral hemisphere in 29.0% of patients in the filter group and in 6.5% of patients in the balloon occlusion group (p = 0.047).
The number of lesions per patient was significantly higher in the filter compared with the balloon occlusion group: 3.6 ± 3.2 versus 1.0 ± 1.4, p = 0.0001 (median [range]: 2 [0 to 13] vs. 0 [0 to 13]) (Fig. 3).
The volume of lesions per patient also differed significantly (0.59 ± 0.6 cm3 for the filter group vs. 0.16 ± 0.2 cm3 for the balloon occlusion group; p = 0.0001), (median [range]: 0.47 [0 to 2.4] cm3 vs. 0 [0 to 0.84] cm3) (Fig. 4).
The 30-day MACCE rate was 3.2% (n = 1 with minor stroke, National Institutes of Health Stroke Scale of 1) in the filter group and 0% in the balloon occlusion group (not significant).
Bleeding or vascular complications were not observed in any patient.
One patient in the filter group crossed over to the balloon occlusion group because of difficulty in placing the filter. One patient in the balloon occlusion group crossed over to the filter group because the stiff wire could not be advanced into the external carotid artery.
Results of patients in the registry
Demographic and clinical characteristics of the patients entered into a registry, including age, percentage of patients age ≥80 years, and symptom status were not significantly different from randomized patients. The incidence of bilateral CAS was significantly higher in patients in the registry compared with randomized patients (63.6% vs. 27.3%, p = 0.002).
Of 22 patients, 20 underwent CAS under filter protection, 1 patient under balloon occlusion, and in 1 patient, no protection system was used. In 14 filter group patients, a DW-MRI could be performed, revealing new ischemic lesions in 85.7% (12 of 14), which is comparable to the group of patients randomized to filter protection (87.1%). One of the 22 patients experienced a minor stroke (4.5%).
The major findings of the present study are:
1. In patients undergoing carotid artery stenting, proximal balloon occlusion significantly reduces the incidence of new cerebral ischemic lesions compared with filter protection as assessed by DW-MRI.
2. Also, the number and the volume of ischemic lesions are significantly lower with proximal occlusion.
3. Eighty-three percent of patients who are candidates for CAS are suitable for proximal balloon occlusion.
To the best of our knowledge, this is the first randomized trial comparing the effectiveness of cerebral embolic protection of a filter versus a proximal balloon occlusion device in all-comer patients undergoing CAS. Cerebral embolization was assessed by DW-MRI, which has been shown to be a sensitive tool in identifying new ischemic cerebral lesions caused by emboli during CAS (11–13). Cerebral ischemic lesions can be detected by DW-MRI in patients undergoing cerebral angiography and CAS in up to 70% of patients (16–18). This is well in accordance with the incidence of new lesions in the present study.
To keep any potential bias as low as possible, the procedure was performed by the same experienced operator, by using the same stent, the same filter, and the same proximal protection device. Variations of these factors could impact the results (14,19). A significant reduction in the incidence, the number, and the volume of new cerebral ischemic lesions was found in patients who underwent CAS with proximal balloon compared with filter protection. This difference was independent of the symptom status of the patients.
A recent study (20) supports our findings. In a subset of patients with high-risk lipid-rich plaque, significantly lower microembolic signal counts as assessed by transcranial Doppler were found in the group of patients undergoing CAS with proximal protection compared with filter protection (20).
Possible explanations for the differences in the efficacy of embolic protection
Compared with balloon occlusion, filters have some disadvantages. They have to cross the lesion before the protection is installed, allow particles smaller than their pore sizes (100 μm) to pass through them or, if not well adapted to the vessel wall, to pass along the filter. Beyond that, filters can become overloaded with debris, with the risk of spilling the contents of the filter during retrieval, and they can occasionally be difficult to retrieve (21,22).
A proximal protection system may be more effective in preventing embolization, mainly because it is placed and functioning before the lesion is crossed, and induces reversed flow by occluding both the external carotid artery and the CCA. In contrast to filters, cerebral protection is kept in place as long as the aspirated blood, which is flushed through a microporous filter, shows visible debris. In multicenter registries, they result in 30-day rates of major adverse events as low as 0.9% to 1.4% (23,24). Nonrandomized comparisons using transcranial Doppler or DW-MRI were either in favor for proximal protection (25) or inconclusive (26,27). Randomized head-to-head comparisons of filters versus proximal balloon occlusion have not been performed so far.
Contralateral new ischemic lesions
They most probably originate from the aortic arch. The carotid artery catheterization technique has an impact on the incidence of new ischemic lesions in the contralateral hemisphere (28).
A lower number of contralateral lesions was found with proximal balloon despite a more complex aortic arch, perhaps as a result of a different approach to the target carotid artery compared with a filter intervention. The external carotid artery was wired first and then a dedicated sheath, which has a very soft tip and a smooth transition to the bigger proximal part, was advanced over this wire, avoiding manipulations in the aortic arch. By contrast, for filter positioning, a 5-F diagnostic catheter was loaded into a 6-F long sheath and advanced into the target artery, followed by the sheath, which then has to cross the aortic arch. This may result in liberating more debris from the aortic arch.
Balloon intolerance as a potential drawback has been described elsewhere (29). Although contralateral occlusion is not a generally accepted contraindication to balloon occlusion, we excluded those patients from the study in order to keep the incidence of balloon intolerance as low as possible. Despite this precaution, balloon intolerance was observed in 13% of the patients in the present study. Other known reasons for balloon intolerance are incomplete or aberrant circle of Willis, simultaneous intracerebral artery occlusive lesions, or poor collateral flow. Pre-procedural imaging by MRI or transcranial Doppler can be useful in identifying these patients; however, no study has proven to be completely reliable in predicting balloon intolerance. Also, in the present study, pre-procedural MRI could not identify all patients who could not tolerate balloon occlusion.
Patients who did not meet the inclusion criteria for the study went into the registry, thus the procedure and outcome data of all consecutive patients were collected. This provides the opportunity to analyze the suitability for filter versus proximal balloon protection in all-comer patients undergoing CAS. All patients in the registry were eligible for filter protection, whereas 15 of 22 patients, i.e., 17%, of the total patient cohort were not suitable for proximal balloon protection. The incidence of new lesions in patients in the registry who received a filter was comparable to the incidence of new lesions of the patients randomized to the filter group, suggesting that the randomized patients were not a highly selected group.
This is a single-center study with a limited number of patients. The endpoint of the study is a surrogate parameter instead of a clinical event. There are differences in baseline patient (presence of coronary artery disease) and lesion characteristics (ulcerated stenosis) that may have introduced bias in favor of proximal protection, yet the differences were not statistically significant because of the small population sample. The extent of coronary artery disease may be a marker for generalized atherosclerosis, which may predispose patients to embolization from the arch. The complex type of the aortic arch, however, was more often found in the balloon occlusion group.
Although the DW-MRI findings were mostly clinically unapparent, they are the tip of the iceberg of what is happening to the brain during CAS. In the present study, proximal balloon protection reduced the incidence of new ischemic lesions by a factor of 1.9. Whether the findings of the present study translate to a lower stroke rate remains to be shown in a randomized trial with clinical endpoints.
In this randomized trial of patients undergoing carotid artery stenting proximal balloon occlusion compared with filter protection significantly reduces the embolic load to the brain.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- carotid artery stenting
- common carotid artery
- carotid endarterectomy
- diffusion-weighted magnetic resonance imaging
- major adverse cardiovascular and cerebral events
- Received September 19, 2011.
- Revision received November 23, 2011.
- Accepted November 29, 2011.
- American College of Cardiology Foundation
- Lloyd-Jones D.,
- Adams R.J.,
- Brown T.M.,
- et al.
- Petty G.W.,
- Brown R.D.,
- Whisnant J.P.,
- Sicks J.D.,
- O'Fallon W.M.,
- Wiebers D.O.
- Adams R.J.,
- Albers G.,
- Alberts M.J.,
- et al.
- Bates E.R.,
- Babb J.D.,
- Caey D.E. Jr..,
- et al.
- Kastrup A.,
- Groschel K.,
- Krapf H.,
- et al.
- Ederle J.,
- Dobson J.,
- Featherstone R.L.,
- et al.,
- International Carotid Stenting Study Investigators
- Mas J.L.,
- Chatellier G.,
- Beyssen B.
- Schlüter M.,
- Tübler T.,
- Steffens J.C.,
- Mathey D.G.,
- Schofer J.
- Jaeger H.J.,
- Mathias K.D.,
- Hauth E.,
- et al.
- Rao A.K.,
- Pratt C.,
- Berke A.,
- et al.
- Montorsi P.,
- Caputi L.,
- Galli S.,
- et al.
- Casserly I.P.,
- Abou-Chebl A.,
- Fathi R.B.,
- et al.
- Ansel G.M.,
- Hopkins L.N.,
- Jaff M.R.,
- et al.,
- Investigators for the ARMOUR Pivotal trial
- Stabile E.,
- Salemme L.,
- Sorropago G.,
- et al.
- Schmidt A.,
- Diederich K.W.,
- Scheinert D.,
- et al.
- El-Koussy M.,
- Schroth G.,
- Do D.D.,
- et al.
- Kim H.J.,
- Lee H.J.,
- Yang J.H.,
- Yeo I.S.,
- et al.
- White C.J.