Benefits of Cerebral Protection During Carotid Stenting With the PercuSurge GuardWire System: Mid-Term Results

ABSTRACT: Purpose. To examine the possible beneficial effect of a new cerebral protection device based on balloon occlusion of the distal internal carotid artery (ICA) and debris aspiration for patients undergoing carotid artery stenting (CAS). Methods. One hundred and eighty-four CAS procedures were attempted under cerebral protection using the PercuSurge system in 167 patients (129 men; mean age 70.5 ± 9.2 years; range 40–91). The lesions were mainly atherosclerotic, and half (n = 93) were asymptomatic. Eighteen restenotic and 7 post-radiation stenoses were also treated. Results. Technical success was 99.5%. All lesions were stented except 3 postangioplasty restenoses. Prophylactic occlusion during protection balloon dilation and stenting was well tolerated in 176 (95.7%) patients. Microscopic analysis of the aspirated blood showed different types of particles numbering between 7 and 145 per procedure with a mean diameter of 250 µm (56–2652 µm). The 30-day stroke and death rate was 2.7%; three periprocedural complications at Cerebrovascular accidents occur each year in 500,000 Americans and result in 150,000 deaths and substantial morbidity.¹ Although antiplatelet agents have a continuing role in reducing cerebrovascular risk, randomized controlled trials have shown that carotid endarterectomy (CE) is superior to medical therapy alone in preventing stroke.^2–4 Carotid endarterectomy, however, has certain limitations. In the North American Symptomatic Carotid Endarterectomy Trial (NASCET), 5.8% of patients had perioperative stroke or death.² In the Asymptomatic Carotid Atherosclerosis Study (ACAS), the perioperative stroke rate was 2.8%.⁴ In higher-risk patients, particularly those with severe coronary artery disease, perioperative morbidity and mortality have been reported in up to 18% of patients.^1,5–14 Cranial nerve palsies occur in up to 27% of patients.^1,10 Also, restenosis develops in 5–19% of cases, and scarring from the initial operation can make repeat surgery more difficult.^9,15 Independent predictors of adverse outcome include contralateral occlusion, previous ipsilateral CE and combined coronary and carotid artery disease.^2,5–14 Further, CE treats only the cervical portion of the carotid artery. Based on the encouraging results with percutaneous interventional techniques in the coronary and peripheral circulations, a natural evolutionary step was the application of these procedures at the cerebrovascular level. Several recent studies suggest that carotid angioplasty and stenting (CAS) can be performed with a perioperative combined stroke and death rate of 2.9–8.2 %.16–22 Shawl et al.²¹ recently showed the safety and efficacy of elective carotid artery stenting in a series of 170 high-risk patients (192 carotid arteries stented). The total 30-day stroke rate was of 2.9% for treated patients or 2.6% for treated arteries. CAS may prove to be safer, less traumatic and more cost-effective than carotid endarterectomy and is not limited to the cervical portion of the carotid artery. Moreover, the risk/benefit ratio may be the greatest in patients at the highest risk for CE.^5,6,12–14 However, embolic stroke, even with a meticulous technique and experienced operators, represents the major drawback of the stenting procedure. The majority of the neurological complications are due to the intracerebral embolization of plaque fragments or thrombus during various stages in the procedures. Cerebral protection devices have been developed to reduce the incidence of embolic events during carotid angioplasty.^23,24 In this prospective study, we examined the outcome of carotid stenting with protection supplied by a distal balloon occlusion device to assess whether this therapy is comparable to historical controls of both carotid endarterectomy and angioplasty without cerebral protection. Methods Study patients.Patients were eligible for this prospective of protected carotid stenting if they had a >= 70% diameter stenosis of the internal carotid artery (ICA) documented by intra-arterial digital substraction angiography and evaluated according to the NASCET criteria.² Patients were excluded for multiple ICA stenoses, intracranial pathology, suspicion of thrombus (e.g., filling defects on the angiogram), gastrointestinal inside the artery at the angiogram in 3 patients), gastrointestinal bleeding in the last 6 months, hemorrhagic disorders, participation in another study during the last 3 months, and inability to give informed consent. Between February 1998 and September 2000, 167 patients (129 men; mean age 70.5 ± 9.2 years, range 40–91) met the inclusion criteria, signed an informed consent statement, and agreed to regular follow-up. Ninety-three (50.5%) patients were asymptomatic (Table 1). The majority had hypertension (125, 74.8%) and over half were smokers (100, 59.9%). Nearly a third of the patients had contralateral disease (36 stenoses and 11 occlusion), and 24 (14.4%) patients had a previously stented contralateral ICA. Owing to comorbities, 122 (73%) patients would have been excluded based on the NASCET or ACAS entry criteria. Of the 184 lesions (17 bilateral) in these patients (Table 2), 157 (85.3%) were atherosclerotic, 18 (9.8%) were restenoses (15 postsurgical, 3 postangioplasty), and 7 (3.8%) were post-radiation stenoses. One lesion was an inflammatory arteritis and another was a post-traumatic aneurysm. The mean percentage stenosis was 81.5% ± 9.3% (range 70–99). Mean lesion length was 14.5 ± 6.2 mm (5–50), and the mean arterial diameter was 5.0 ± 1.2 mm (4.1–7.1). Based on duplex scans, 85 (46.2%) lesions were calcified, 135 (73.4%) were ulcerated, 87 (47.3%) were hyperechogenic, and 96 (52.2%) echolucent. Devices and stenting technique. A variety of stents were implanted in this study, including the Palmaz (Cordis Europe, Roden, The Netherlands), Corinthian (Cordis), Wallstent (Boston Scientific Corporation, Natick, Massachusetts), nitinol self-explanding stents (Bolton Medical, Villers-les-Nancy, France), and Jostent (Jomed AB, Helsingborg, Sweden). The GuardWire System (Medtronic AVE/PercuSurge, Sunnyvale, California) has been described previously,²⁴ but briefly consists of: • A 0.014" or 0.018" hollow nitinol wire incorporating into its distal segment an inflatable, compliant balloon that is available in 3–6 mm diameters. • A Microseal incorporated at the proximal end of the wire to facilitate balloon inflation and deflation via a Microseal adapter; the Microseal keeps the balloon inflated while allowing catheter exchange at the proximal end similar to commonly used guidewires. • An over the wire aspiration catheter to remove debris and flush the ICA. Prior to the procedure, patients had been given ticlopidine (250–500 mg/day) or clopidogrel (75 mg/day) for at least 2 days and preferably 1 week. In the catheterization laboratory, the percutaneous procedures were approached through the common femoral artery. Unfractionated heparin (5,000 units) and atropine (1 mg) were routinely administered intravenously just after introducer sheath placement. A temporary pacemaker was never used. An 8 or 9 Fr multipurpose guide catheter was delivered into the common carotid artery (CCA). The GuardWire was then gently advanced through the guide catheter until the marker of the protection balloon had been placed 2 or 3 cm beyond the lesion (Figure 1). The MicroSeal adapter was then attached, and the protection balloon slowly inflated with a fixed volume of dilute contrast, occluding the ICA and diverting vessel outflow toward the external carotid artery (ECA). On detaching the Microseal adapter, the occlusion balloon remains inflated. Predilation of the lesion or direct stenting were then performed under protection using either of two techniques. In the first, the protection balloon remained inflated during the entire procedure, and aspiration was performed once after stent placement and post-dilatation. In the alternative method, the occluding balloon was deflated between predilation and stent placement to restore cerebral flow, with aspiration performed after each stage. The technique used depended on the patient’s tolerance to occlusion, the cerebral collateral circulation, the status of the contralateral artery, the duration of the procedure, and the technical problems encountered. In both scenarios, the aspiration catheter was advanced over the wire into the dilated area; debris was removed to a 20-mL syringe connected to the catheter. In our initial 40 cases, the treated area was then flushed to “drive” the particles towards the ECA. The saline injections were delivered through the guide catheter using an injection pump at a rate of 2 ml/s for 10 seconds. The aspirated blood samples were collected in 40-mm filters and sent for analysis using optical and electron microscopic techniques. Finally, the Microseal adapter is reattached to the GuardWire, the P.B. deflated allowing normal flow to restore. If the angiographic result is satisfactory, the device is removed. After aspiration, the Microseal adapter was reattached to the GuardWire, and the protection balloon was deflated, restoring normal flow. If the angiographic result was satisfactory, the device was retrieved and the procedure completed. The introducer sheaths were removed when the acitvated coagulation time (ACT) was Periprocedural assessment. A neurological assessment by an independent neurologist using the N-score clinical examination scale²⁵ was routinely performed at baseline and following the procedure, as were blood chemistries, coagulation profile, electrocardiogram, duplex studies, and computed tomography (CT) or magnetic resonance imaging (MRI) of the brain. Transcranial Doppler was done before and after the procedure but not intraprocedurally as a rule; neurological status was continuously monitored by contralateral hand-gripping maneuvers during balloon inflation and after stenting. During the duplex study, plaque echodensity was characterized in a subgroup of 20 patients according to the Nicolaides’ method,^26–27 which represents plaque echogenicity as a gray scale median density value (MDV). The MDV was compared to the number and size of the particles released during CAS and the relationship between the degree of stenosis (represented by the peak systolic velocity) and the number of particles. Computerized digital subtraction angiography included at least two orthogonal projections of the carotid bifurcation and images of the intracranial circulation before and after the procedure. The NASCET angiographic criteria² were used to calculate the degree of stenosis before and after angioplasty and stenting, with the distal, nontapering portion of the ICA serving as the reference segment. Surveillance. All patients underwent a neurological examination and duplex CT scans the day after CAS. At 30 days and every 6 months thereafter, the neurological examination and duplex scan were repeated. Angiography was performed at 6 months. Any change in neurological status required repeated CT brain scan, which was read by an independent observer blinded to the patient outcome. Definitions, endpoints, and analyses. Among the complications, myocardial infarction (MI) referred to the development of new Q waves on the electrocardiogram (ECG) and/or a creatine kinase elevation to at least twice the normal level, accompanied by above-normal elevation of the MB band. Transient ischemic attacks (TIA) were new neurological deficits that completely resolved within 24 hours. Minor strokes were new neurological deficits that persisted for > 24 hours but completely resolved or returned to baseline within 7 days, thus, they were non-disabling neurological events. Major strokes, on the other hand, were new neurological deficits that persisted after 7 days. The primary clinical endpoints included any major/minor stroke, death or MI within the first 30 days after the procedure; any of these sequelae occurring within 48 hours of the procedure constituted the periprocedural complications. The secondary clinical endpoints were the need for new intervention, angioplasty or endarterectomy at 6 months. Angiographic success was defined as achieving a 50%. Procedural success was defined as a reduction in the stenosis to Tolerance to occlusion balloon. Prophylactic occlusion during balloon inflation was well tolerated in 176 (95.7%) lesions, 47 of which were in patients with significant contralateral ICA disease (stenosis or occlusion). Two (1.1%) patients became completely intolerant to protection balloon inflation. One patient, who had total occlusion of the contralateral ICA, lost consciousness and developed seizures. The patient recovered fully after rapid balloon deflation. CAS was successfully completed without cerebral protection. The other patient had poor collateral circulation via the circle of Willis and lost consciousness rapidly upon balloon inflation, but the procedure was completed under protection. The patient recovered rapidly after deflation of the protection balloon. Six (3.3%) patients also demonstrated partial transient intolerance beginning approximately 2 minutes after flow interruption with transient symptoms such as agitation, brief loss of consciousness, or transient neurological deficit. The procedure was completed under protection, and all patients had rapid and complete recovery while the protection balloon was still inflated. Four of them had a brief hypotensive response to dilation with bradycardia, which may have been responsible for the intolerance. Six patients developed a spasm of the ICA above the dilated area at the location of the protection balloon; this rapidly responded to vasodilator therapy. Debris and plaque echogenicity analyses. Visible debris (mean diameter 250 µm; range 56–2,652) was extracted from all patients at a mean of 74 particles per procedure (range 7–145). Different types of particles were found: atheromatous plaques, cholesterol crystals, calcified crystals, necrotic cores, fibrin, recent and old thrombi, platelets, macrophage foam cells, lipoid masses and acellular material. In a subgroup correlation of particulates to the plaque’s MDVs (Figure 2), hypoechoic plaques had low MDVs and hyperechoic plaques were high. The hypoechoic plaques were correlated to a greater number of particles (r2 = 0.72). There was also good correlation between the MDV and the mean size of the debris (r2 = 0.75), indicating that hyperechoic plaques produced larger debris. No relationship was found between the degree of stenosis and the number of particles. Complications. There was 1 (0.6%) death of a symptomatic patient from cardiac failure 3 weeks after the CAS procedure. Four (2.2%) neurological complications associated with the 184 procedures occurred within 30 days, three (1.6%) periprocedurally (1 amaurosis and 3 TIAs), for a 30-day stroke and death rate of 2.7%. In the first of the periprocedural events, a patient with a high-grade, ulcerated ICA stenosis developed amaurosis after acute intraprocedural thrombosis of a Wallstent. The thrombosis was seen on the angiogram after deflation of the protection balloon, which was quickly re-inflated. Abciximab was administered, and 10 minutes after the bolus, thromboaspiration and flushing through the guide catheter were performed. The protection balloon was finally deflated, and the final angiogram showed no residual thrombus inside the stent. However, the patient developed amaurosis, which was believed at the time to have arisen from an embolism reaching the ophtalmic circulation via the ECA. Indeed, such a communication was diagnosed after careful angiographic inspection. The 2 TIAs manifested as a unilateral upper arm paresis in a symptomatic patient and as transient hemiparesis in an asymptomatic patient who had prolonged occlusion time (19 minutes) and a tight ICA stenosis. No evidence of ischemia was detected at subsequent serial CT examinations. The fourth complication occurred in one of the symptomatic patients who had received intraprocedural abciximab for subtotal occlusion. On the third day after stenting, he developed an intracerebral hemorrhage with hemiparesis; he was partially recovered 2 months later. Follow-up. The mean follow-up was 335 ± 165 days (range 1–31), during which 3 (1.8%) patients died, two from MIs and 1 from a major stroke at 6 months after stenting in the contralateral ICA. No other minor or major stroke occurred. One (0.5%) asymptomatic restenosis was observed at 6 months, which was treated successfully by repeat angioplasty. The event-free survival was 97% at 20 months (Figure 3). Discussion Recent randomized trials have proved the efficacy of surgical endarterectomy for severe symptomatic and asymptomatic extracranial carotid artery stenosis and its superiority over medical treatment.^2–4,28,29 However, the benefits of the procedure are critically dependent on the rate of perioperative complications.^10,12,13 To be considered as an alternative to surgery, CAS must demonstrate similar complication rates. CAS has been offered to an increasing number of patients with carotid artery stenosis based on what appears to be acceptable perioperative stroke/death rates.^{16–23,30–34} However, even in the hands of experienced interventionalists, the risk of embolic stroke remains the main limitation of the procedure. Several years ago, Ohki et al.³⁵ demonstrated the frequency of debris migration and distal embolism in an ex vivo human carotid stent model; their observatons have since been confirmed clinically.^12,36 While the number of embolic particles generated by percutaneous techniques seems far in excess of those associated with endarterectomy,^33,35,36 their clinical significance has not been documented yet,^36,37 nor has the minimum particle size capable of producing ischemic events. Undoubtedly, though, the presence of these particulates cannot have any beneficial effect on the brain. Various patient and plaque characteristics have been suggested as predictors of debris generation and embolic events^35,38 to define high-risk groups for CAS procedure. In our study, debris were extracted from all patients, even in lesions theoretically at low risk of cerebral embolism (restenosis, echogenic plaques, concentric lesions) suggesting that the risk of embolization is independent of the nature of the plaques. Additionally, stent deployment does not provide sufficient protection against plaque embolization. In all CAS series, the embolic risk exists regardless of the implantation technique and the stent characteristics. Manninen et al.³⁸ compared endovascular stent placement with balloon dilation of carotid arteries in cadavers in situ and found no difference with respect to distal embolization. In 1984, Vitek and colleagues³⁹ first reported a case of successful innominate artery angioplasty, in which the risk of cerebral embolization was reduced by temporary balloon occlusion at the origin of the right CCA. Over the last decade, the need to overcome suboptimal results and eliminate embolic risk has produced several suggestions for protection techniques during carotid angioplasty.^40,41 The PercuSurge Guardwire device was first tested in animals by Oesterle et al.,⁴² followed by clinical use in 27 coronary saphenous vein graft angioplasties.⁴³ The system has been shown to be compatible with routine angioplasty procedures and capable of containing and retrieving atherosclerotic debris that might aid in the prevention of distal embolization during carotid angioplasty. One of this system’s advantages is its mimicry of steerable coronary guidewires, allowing easy navigation through stenoses with minimal technical failures. We experienced only one failure to cross a tight stenosis in a tortuous carotid artery. Additionally, the GuardWire provides both sufficient support to advance the dilation balloon and stent and rapid deflation time of the occlusion balloon (~ 15 seconds). Limitation of the techniques. This study showed that protected CAS is a feasible and safe procedure with a very low (2.2 %) 30-day neurological complications rate, which is comparable to other studies of unprotected carotid stenting and historical surgical controls.^{16,18,21,37,44,45} However, cerebral protection cannot prevent all embolic events that may occur at all steps in the procedure. The balloon protection device is active against embolism only after the wire has crossed the lesion; this maneuver and even the initial positioning of the guide catheter in the CCA can release embolic material. Utilization of smaller tools and adaptation of coronary techniques may limit these risks and improve outcome in this regard. The issue of patient tolerance of balloon occlusion is important. Before the procedure, complete angiographic assessment of the 4 supra-aortic vessels is mandatory for determination of the vertebrobasilar and contralateral carotid flows and the collateral supply through the circle of Willis. Patients with congenital absence or acquired disease of these structures may not tolerate flow occlusion. This problem is similar but not identical with surgical clamping during carotid endarterectomy, since flow through the ECA is unaffected with protection balloon inflation. The ECA also provides, through collaterals, flow to both the anterior and posterior cerebral circulation, which is useful when the ICA is occluded, but potentially harmful if flushing is used to clean the protected vessel segment. In our study, ICA occlusion was well tolerated in the majority of cases. Only one patient demonstrated extreme intolerance, causing us to complete the procedure without protection. The other 7 instances were the more common delayed intolerance of brief duration, which began after the procedure was well advanced, usually after stent deployment and before debris aspiration. In these cases, the procedure could be completed with aspiration and reestablishment of the cerebral flow, thus preserving the protective benefits. After aspiration, potential remaining debris may be flushed toward the ECA and lead to ischemic complications in cases of distal anastomosis between the ECA and the ICA or vertebrobasilar artery territory through the meningeal or occipital arteries. Diagnostic angiography prior to treatment is mandatory for diagnosis of these aberrancies, which are more prevalent than expected. Moreover, even if there is no angiographic evidence of a communication between the ECA to intracerebral vasculature, this does not preclude the opening of this collateral route during ICA occlusion. In our series, one neurological complication appeared after flushing, which caused us to abandon this flushing technique. We now firmly believe that debris removal must be restricted to aspiration. Alternative approaches to cerebral protection. A variety of cerebral protection devices and techniques have been proposed recently. Kachel40 used a proximal balloon occlusion technique, which consisted of balloon occlusion of the upper part of the CCA during the carotid angioplasty. Filters are another viable option for cerebral protection. The main advantage of filters is their ability to function in line with the flow; unfortunately, their large pores (>= 100 µm today) allows microparticles through,⁴⁷ which could lead to ischemic, particularly ocular, complications and cognitive changes. However, filters may have a more specific indication in patients with insufficient cerebral supply who would barely be able to tolerate balloon occlusion. Nevertheless, in our series, we have seen that even with a contralateral carotid disease, the occlusion balloon was well tolerated in most of the cases. The reversal of flow technique recently proposed by Parodi et al.⁴⁸ with its occlusion of the ECA and upper CCA seems very promising.⁴⁹ This technique allows the lesion to be crossed under protection, avoiding cerebral embolism during this step. Over a hundred patients have been treated with this technique, and no related-device neurological embolic events have occurred.⁵⁰Study limitations. This was a prospective, nonrandomized, single-center study in which CAS under cerebral protection was performed by highly experienced interventionalists; whether similar results will be obtained by less experienced operators is not known. Moreover, this study represents early clinical experience with equipment developed for coronary and peripheral vascular interventions. Devices designed specifically for carotid artery intervention may improve outcome. Clinical implications and futures studies. In the low-risk patients randomized into NASCET and ACAS, relief of obstruction was shown to lower the risk of cerebrovascular events. Whether other patient groups with different baseline characteristics would have the same treatment advantage is not known with certainty, nor is the relative effectiveness of CAS and CE in preventing ischemic brain infarction and death in high-risk patients. Shawl et al.²¹ saw very few neurologic events during the 19-month follow-up of their patients, suggesting that the effectiveness of recanalization may well be reflected in long-term clinical benefit. The results of our CAS under cerebral protection series are similar and very promising. For this reason, randomized controlled trials of CE versus CAS are the next step in evaluating CAS. Until the results of the recently begun randomized trials are available, caution should be exercised in discarding CE in patient groups with proven efficacy. One randomized trial, the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS), which examined the role of angioplasty versus CE, has been completed.⁵¹ This trial, although underpowered, suggested that balloon angioplasty without routine stenting has a similar safety profile to elective CE. These data suggest that routine stent implantation will further improve the percutaneous management of carotid artery disease. A second trial that compares CE and CAS, the Carotid Revascularization Endarterectomy Trial (CREST), will commence early in 2002,⁵² and the final results will not be available for at least 5 to 6 years. In the interim, there are sufficient published reports to support the use of CAS by experienced operators in patients at high surgical risk: those with carotid artery lesions above the C2 or C3 cervical vertebra or at the CCA ostium, and patients with cervical spine disease or fixation, previous radical neck dissection, fibromuscular dysplasia, previous cervical radiation, previous CE, and important comorbid conditions, including unstable angina, recent MI and severe congestive heart failure. In addition, there will be a continuing evolution of new stents, dilation and postdilation strategies, distal neuroprotection devices, all of which will require evaluation. Moreover, the cost of these different techniques has to be evaluated. In conclusion, carotid artery stenting has been demonstrated as feasible and safe, even in high-risk patients with a complication rate comparable to that of patients in the ACAS and NASCET trials. However, CAS without cerebral protection is associated with the risk of brain embolism. The addition of a protection device may make the complication rates of CAS equal to or even less than those obtained with carotid endarterectomy.