Historical Perspective

Mathias and colleagues performed the first reported angioplasty of a carotid bifurcation in 1980.⁸ The indications initially were mostly nonatherosclerotic disease (radiation-induced or inflammatory stenosis); however, soon there was a push toward atherosclerotic disease, and that resulted in the discovery of an inordinate risk for iatrogenic dissection (5% to 8%) and distal embolic complications (8% to 10%).^9,10 This initial finding resulted in the impetus for discovery of a distal embolic shower protection (DEP) device. The first attempts were simply distal occlusion of the internal carotid artery through a balloon parked in parallel to the more proximal stent delivery system, followed by aspiration of blood and debris after angioplasty. This was refined by the development of a wire-mounted balloon (PercuSurge GuardWire; Medtronic AVE/PercuSurge Inc., Sunnyvale, CA) that allowed a balloon to be passed through the lesion for angioplasty while there was distal flow arrest. However, the results from carotid angioplasty remained dismal, principally from the high rate of recurrent stenosis, as well as procedural complications. Parallel development of coronary and peripheral vascular stents eventually resulted in the trial of a stent at the carotid bifurcation with inherent jailing of the external carotid artery in 1990.¹¹ This step single-handedly jettisoned CAS into a viable solution initially for patients considered very poor candidates for CEA.

The revolution that ensued has seen a wide variety of balloons for angioplasty, stents designed specifically for the carotid anatomy, an array of distal protection devices, and most recently, an entirely new way of distal protection achieved through flow reversal from the internal carotid artery into the arterial guide sheath, which is the conduit for deployment of devices across the carotid bifurcation. The major impetus for advancement of CAS came with publication of the results of the Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial,¹² which effectively demonstrated that patients considered to be at high risk with CEA were less likely to have complications if treated by CAS. This trial resulted in Food and Drug Administration (FDA), Centers for Medicare & Medicaid Services (CMS), and Medicare approval of CAS as a viable option in such patients. This brings us to the current era, during which a large number of trials that have been completed and others that are ongoing continue to refine our understanding of the uses and restrictions of CAS. The future for CAS appears to be clear: it is here to stay.

Carotid Endarterectomy Trials

Indications for and outcomes of CEA have been extensively studied. Support for performance of CEA is generated from four well-designed multicenter, randomized clinical trials—NASCET,^2,3 ECST,^6,7 ACAS,¹³ and ACST.⁵ NASCET^2,3 and ECST^6,7 addressed the use of CEA for symptomatic patients with 70% to 99% carotid stenosis or selected patients with 50% to 69% stenosis (Table 351-1). These studies resulted in class IA indications for the use of CEA in symptomatic patients meeting the appropriate criteria.¹⁴ However, it is important to realize that the general population of patients with carotid stenosis has substantially different demographics than do those who met the strict eligibility criteria for these studies.¹⁵ For instance, NASCET excluded patients 80 years and older and those with intracranial carotid stenosis more severe than the surgically accessible lesion; liver, kidney, or lung failure; cardiac valve or rhythm disorder; previous ipsilateral CEA; uncontrolled hypertension or diabetes; recent myocardial infarction (MI); or major surgery.² Such patients were considered to have excessive perioperative morbidity (i.e., high risk). Since publication of the NASCET results, patients considered for carotid revascularization are often divided into low- and high-risk groups, and this surgical risk stratification has been applied as an integral part of the study design in recent CAS trials.

The ACAS trial¹³ and the ACST⁵ addressed the use of CEA in asymptomatic patients. The degree of benefit from CEA for asymptomatic lesions is substantially less, and the indications for revascularization are still debated. ACAS and ACST demonstrated a 5.4% to 5.9% absolute reduction in risk over a 5-year period.^5,13 Therefore, periprocedural risks are particularly relevant to the decision analysis for the treatment of asymptomatic patients, with a morbidity rate higher than 3% minimizing any benefit. Nonetheless, as a result of publication of the ACAS trial, nearly 75% of CEAs in the United States are performed on asymptomatic patients.¹⁶

In the aforementioned trials, carefully selected low-risk patients were treated by highly experienced surgeons at high-volume medical centers. The low complication rates seen in NASCET and ACAS are often not obtained in the general population. Studies have shown perioperative stroke and death rates to range from 0%¹⁰ to 11.1%¹⁷ for symptomatic patients and from 0%¹⁸ to 5.5%¹⁷ for asymptomatic patients. In fact, a study of Medicare mortality data from hospitals participating in NASCET and ACAS demonstrated a 1.4% perioperative mortality rate¹⁵ as compared with 0.6% reported in NASCET² and 0.1% reported in ACAS.¹³ Perhaps equally concerning, CEA-related mortality rates have been demonstrated to be higher (2.5%) for low-volume hospitals,¹⁵ although other studies have argued that only small differences exist between mortality rates at high- and low-volume hospitals.¹⁹

Treatment decisions are also dependent on patient-specific factors. The presence of comorbid disease has a significant influence on outcome after CEA. Perioperative stroke and death rates for common comorbid conditions are 8.6% for congestive heart failure,^20,21 7.5% for age older than 75 years,^20,21 10.8% for postendarterectomy restenosis,²² 13.9% for ipsilateral carotid siphon stenosis,²⁰ 10.7% to 17.9% for intraluminal thrombus,^20,23 14.3% for contralateral carotid occlusion,²⁴ and 16.4% to 26.2% for CEA combined with CABG.^25,26 It is important to note that in the presence of such comorbidity, the natural history of carotid disease itself is grimmer. The investigators of the Asymptomatic Carotid Stenosis and Risk of Stroke (ACSRS) “natural history” study monitored 1115 patients with asymptomatic internal carotid artery stenosis treated with medical therapy alone and identified significant differences in patient subgroups with respect to risk for stroke and death.²⁷ The highest risk group (82% to 99% stenosis according to NASCET criteria,² history of contralateral transient ischemic attack [TIA], and serum creatinine level >0.085 mmol/L) had a 4.3% annual ipsilateral stroke rate as opposed to 0.7% in the lowest risk group.^27,28

It should also be noted that since the aforementioned major randomized CEA trials began, best medical therapy has improved. In NASCET, the primary medical intervention was 1300 mg of aspirin on a daily basis.² This dose of aspirin is no longer used because lower doses have been proved to be equally efficacious with fewer side effects.^29–31 Other antiplatelet drugs, such as clopidogrel and ticlopidine, are also now available,^32,33 and the combination of aspirin and dipyridamole was shown to be more efficacious than aspirin alone.³⁴ Methods for blood pressure control were not specified in NASCET, whereas it is now known that blood pressure below 120 to 130/70 mm Hg is optimal for reduction of cardiovascular risk in patients with medical comorbid conditions^14,35,36 and that for primary stroke prevention, a 10–mm Hg reduction in systolic blood pressure produces a 31% reduction in the relative risk for stroke.³⁷ For secondary stroke prevention, angiotensin-converting enzyme (ACE) inhibitors^35,38 and the combination of a thiazide diuretic with an ACE inhibitor³⁸ have now been proved effective. Additionally, in the past decade, statins have assumed a prominent role in cerebrovascular and cardiovascular risk modification.^39–43 In a study of patients receiving medical therapy for severe carotid artery disease, statin use was associated with significantly lower rates of stroke, MI, and death.⁴⁴ It is likely that improvements in medical therapy for carotid atherosclerotic disease and related comorbidity should prompt periodic reevaluation and fine-tuning of the risk-benefit analysis for medical therapy versus surgical intervention.

With the great deal of complexity regarding risk assessment in this complex patient population, current standards are limited to minimizing overall surgical risk to maximize the probable benefit from surgery. Currently, the guidelines of the American Heart Association/American Stroke Assocation¹⁴ and the Canadian Neurosurgical Society⁴⁵ have established an upper limit of 6% for perioperative risk in symptomatic patients¹⁴ and a 3% upper limit in asymptomatic patients, assuming a life expectancy exceeding 5 years.²⁶

Publication of Carotid Angioplasty and Stenting Study Data

The results of trials comparing endovascular and surgical treatment of carotid stenosis are summarized in Table 351-2.^46–56 The first randomized trial comparing these treatments, the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS), included 504 patients enrolled between 1992 and 1997 and was designed to compare balloon angioplasty alone with CEA.⁴⁶ Stents, when they became available, were incorporated as well but accounted for only 26% of cases. Twenty-four centers in Europe, Australia, and Canada participated, and as in previous CEA trials, high-risk surgical patients were excluded from enrollment, including those with recent MI, poorly controlled hypertension or diabetes mellitus, renal disease, respiratory failure, inaccessible carotid stenosis, or severe cervical spondylosis. CAVATAS demonstrated no statistically significant difference between endovascular and surgical treatment in the rates of disabling stroke or death within 30 days (6.4% for CAS versus 5.9% for CEA) and no significant difference in 3-year ipsilateral stroke rates. It is interesting to note that only 26% of patients randomized to endovascular treatment actually received stents. This becomes even more pertinent when one considers the 3-year ipsilateral stroke rates. These early encouraging results generated a great deal of interest in CAS, and further studies were undertaken.

The Wallstent trial,^47,48 the first multicenter randomized trial designed from inception to evaluate CEA and CAS equivalence, enrolled a total of 219 symptomatic patients with 60% to 99% stenosis. Thirty-day stroke or death rates were 12.1% with CAS and 4.5% with CEA (P = .049). Additionally, 12.1% of CAS patients suffered ipsilateral stroke, procedure-related death, or vascular death at 1 year versus 3.6% of CEA patients (P = .022), and as a result, the trial was halted by the Data Safety and Monitoring Committee after an interim analysis demonstrated worse outcomes for the CAS group. Critical to interpreting these results is the fact that distal protection devices were not used in the Wallstent trial. A significant proportion of major CAS neurological complications are due to embolization of atheromatous material.^57–60 Devices that capture embolic debris released during CAS have significantly improved procedural safety.^57,60–64

One of the first trials to use embolic protection was Carotid Revascularization Using Endarterectomy or Stenting Systems (CaRESS),^54,55 a multicenter, nonrandomized, prospective study comparing CAS with embolic protection (n = 143) and CEA (n = 254) in symptomatic (32%) and asymptomatic (68%) patients at low and high surgical risk. An important feature of CaRESS was that the treatment procedure was chosen by the treating physician and the patient, not randomized. Although this study design probably introduced selection bias, the CaRESS trial represents a generalized perspective on carotid revascularization and more closely represents its “real-world” application. Baseline group demographics were similar, except that patients with previous carotid interventions more often underwent CAS. No statistically significant difference between 30-day and 1-year death or stroke rates existed between CAS and CEA (2.1% versus 3.6% and 10.0% versus 13.6%, respectively), nor were significant differences found for restenosis, residual stenosis, repeat angiography, and need for carotid revascularization. Overall morbidity and mortality approached NASCET^2,3 and ACAS¹³ standards and represent the lowest rates among the major CAS trials to date. The low stroke and death rates may be attributable to the ability of the treating physician to consider patient-specific factors and successfully assign each patient to the safest therapy.

CAS was well established as a treatment option for high-risk patients by SAPPHIRE,¹² a randomized, multicenter trial designed to determine CAS noninferiority to CEA in high-risk patients. Eligible patients (n = 334) had symptomatic stenosis of at least 50% or asymptomatic stenosis of at least 80%. The 30-day MI, stroke, or death rate was 4.8% for CAS and 9.8% for CEA (P = .09). Much of this difference was secondary to MIs occurring in the CEA group, and although not reported in the SAPPHIRE publication, the 30-day rate of stroke and death was approximately 4.8% for CAS patients and around 5.6% for CEA patients. At 1 year, 12.2% of CAS patients had suffered stroke, MI, or death versus 20.1% of CEA patients (noninferiority analysis: P = .004; superiority analysis: intention to treat, P = .053; as treated, P = .048). MI and major ipsilateral stroke rates were significantly better after CAS than after CEA (2.5% versus 8.1%, P = .03; 0% versus 3.5%, P = .02; respectively).

Because SAPPHIRE had shown such clear noninferiority in high-risk patients, the Stent-supported Percutaneous Angioplasty of the Carotid artery versus Endarterectomy (SPACE) trial⁵⁰ set out to establish noninferiority for CAS versus CEA in low-risk patients. In this multicenter trial, the safety and efficacy of CAS and CEA were compared in 1183 randomized patients with symptomatic carotid artery stenosis (≥70% by duplex ultrasonography, ≥50% by NASCET criteria,² or ≥70% by ECST criteria⁶). The 30-day rates of ipsilateral stroke or death were 6.84% for CAS and 6.34% for CEA (P value not significant).⁵⁹ It is important to note that embolic protection was not required and was used in just 27% of patients, although a subgroup analysis did not demonstrate a significant difference between patients with embolic protection and those without. Despite these encouraging results, “SPACE failed to prove the non-inferiority of carotid-artery stenting”⁵⁰ statistically. This is because the trial was halted more than 700 patients shy of its goal enrollment of 1900 as a result of an interim analysis demonstrating that 2500 patients would be needed to reach significance given the results up to that point. The steering committee acknowledged a “lack of funds”⁵⁰ to expand the trial to an enrollment of 2500 patients and therefore halted the trial. In essence, the study was underpowered to demonstrate noninferiority because of incorrect estimation of the anticipated effect sizes. Still, although its a priori goals were not realized, the 0.51% difference observed in perioperative stroke or death between CAS and CEA was not statistically significant and is well within the published differences between individuals, institutions, and variations of CEA.

The SPACE results, even though they were negative, were still quite encouraging to CAS proponents. Unfortunately, a second multicenter, randomized trial to assess the noninferiority of CAS versus CEA in patients with more than 60% stenosis, Endarterectomy Versus Angioplasty in Patients with Symptomatic Severe Carotid Stenosis (EVA-3S), was ended after interim analysis (n = 527) demonstrated the 30-day rate of any stroke or death to be significantly higher in the CAS group (9.6%) than in the CEA group (3.9%) (P = .01).⁵² Importantly, early in the trial the use of embolic protection was not required. Patients treated without embolic protection experienced a 25% 30-day rate of stroke or death (5 of 20 patients), which prompted changes in protocol by the EVA-3S safety committee. Additionally, EVA-3S compared groups of physicians with unequal experience. Surgeons performing CEA had performed at least 25 endarterectomies in the year before trial entry, yet endovascular physicians were certified after completing as few as 5 to 12 CAS procedures (5 CAS among at least 35 stent procedures on supra-aortic vessels or 12 CAS). Endovascular physicians were also allowed to enroll study patients while simultaneously undergoing training and certification. Subgroup analysis based on CAS physician experience demonstrated a 12.3% stroke and death rate for endovascular physicians tutored in CAS during the trial,⁵² 7.1% for those tutored in CAS during their endovascular training, and 10.5% for physicians with experience in CAS. The resulting overall rate of stroke and death (9.6%) is substantially higher than that in other randomized trials. Therefore, it is hard to accept such an elevated complication rate as representative of the practice of CAS in general. It is more likely that EVA-3S emphasizes the importance of embolic protection, as well as rigorous training and credentialing for CAS physicians. The implied importance of embolic protection in EVA-3S has been further supported by numerous imaging studies examining the frequency of (mostly small, asymptomatic) ischemic lesions on postoperative magnetic resonance imaging (diffusion-weighted imaging [DWI]). These studies have demonstrated the following: a reduction in the frequency of lesions seen on DWI with distal embolic protection (49% versus 67%)⁶² and fewer lesions noted on DWI after CEA than after CAS (11.6% versus 42.6%, no significant clinical difference) with current embolic protection devices,⁶⁵ as well as a low frequency of DWI-confirmed lesions with more recent embolic protection devices, such as the Parodi flow reversal system (NeuroProtection System, W. L. Gore & Associates, Flagstaff, AZ),⁶⁶ at a rate not significantly different from that incurred by diagnostic cerebral angiography alone (18.2% versus 11.5%).⁶⁷

Carotid registries are nonrandomized outcome records for symptomatic and asymptomatic high-risk CAS patients. These registries include Arbeitsgemeinschaft Leitende Kardiologische Krankenhausarzte (ALKK); ACCULINK for Revascularization of Carotids in High-Risk patients (ARCHeR); Boston Scientific EPI: A Carotid Stenting Trial for High-Risk Surgical Patients (BEACH); Carotid Artery Revascularization using the Boston Scientific FilterWire EX/EZ (CABERNET); Carotid Acculink/Accunet Post Approval Trial to Uncover Unanticipated or Rare Events (CAPTURE); Carotid Artery Stenting with Emboli protection Surveillance—Post Marketing Study (CASES-PMS); and Carotid Revascularization with ev3 Arterial Technology Evolution (CREATE) (Table 351-3).^68–75 Although registries do not provide direct comparison data, they do help establish true adverse event rates in high-risk CAS patients and are a crucial component for improving our understanding of the risks associated with CAS. The collaborators of CABERNET found a 4.0% 30-day rate of death, stroke, and MI (N = 446 patients),⁷¹ whereas the investigators of ARCHeR (N = 581 patients) found a 30-day stroke or death rate of 6.9%, as well as a 1-year composite outcome (30-day rate of MI, stroke, or death plus the 1-year rate of ipsilateral stroke) of 9.6%.⁶⁹ CREATE (N = 419 patients) demonstrated a 6.2% 30-day rate of MI, stroke, and death.⁷⁵ The CAPTURE registry (N = 3500) determined that the post-CAS incidence of stroke, MI, and death was 6.3% for patients treated with the Acculink/AccuNet CAS system (Abbott Vascular, Santa Clara, CA), as well as a rate of major stroke or death of 2.9%.^72,73 The BEACH investigators (N = 747 patients) found a 30-day MI, stroke, or death rate of 5.8%.⁷⁰ These results were similar to those in the CASES-PMS registry (5.0%), which examined the use of distal protection by endovascular carotid surgeons who either had previous experience with the device (AngioGuard XP, Cordis Endovascular, Warren, NJ) or underwent formal training (N = 1493).⁷⁴ Under these rigorous conditions, the 30-day major adverse event rate did not vary significantly between symptomatic and asymptomatic patients and among physicians with high and low volume or differing level of experience with the specific distal protection device. The German ALKK registry (N = 1888 patients), which included patients with standard risk, demonstrated an in-hospital death and stroke rate of 3.8%.⁶⁸ Interestingly, when this risk was stratified by time, the investigators saw improvement from 6.3% in 1996 to 1.9% in 2004 (P = .021). Continued effort to maintain rigorous registries such as those just described is critical to our eventual understanding of appropriate patient selection and procedural risks.

TABLE 351-3 Carotid Angioplasty and Stenting Registries

ALKK, Arbeitsgemeinschaft Leitende Kardiologische Krankenhausarzte; ARCHeR, ACCULINK for Revascularization of Carotids in High-Risk patients; BEACH, Boston Scientific EPI: A Carotid Stenting Trial for High-Risk Surgical Patients; CABERNET, Carotid Artery Revascularization using the Boston Scientific FilterWire EX/EZ; CAPTURE, Carotid Acculink/Accunet Post Approval Trial to Uncover Unanticipated or Rare Events; CASES-PMS, Carotid Artery Stenting with Emboli protection Surveillance—Post Marketing Study; CREATE, Carotid Revascularization with ev3 Arterial Technology Evolution; MI, myocardial infarction.

Current Studies

The two major most recently completed randomized trials of CAS versus CEA are the Carotid Revascularization Endarterectomy versus Stent Trial (CREST) and the International Carotid Stenting Study (ICSS). CREST was a National Institutes of Health–funded, multicenter randomized trial that enrolled 2502 patients with greater than 50% symptomatic carotid stenosis or greater than 70% asymptomatic stenosis for randomization to CEA or CAS.^76–80 Primary end points included death, stroke, or MI at 30 days and ipsilateral stroke within 60 days. CREST maintained a rigorous credentialing phase for CAS providers. CREST demonstrated no significant difference in the primary endpoint (30-day MI, stroke and death rate) between CAS (5.2%) as compared with CEA (4.5%) (p = 0.38). The major differences between the two treatments were that minor strokes (not major) were more common following CAS (3.2 vs. 1.7%, p = 0.01) whereas MI (1.1 vs. 2.3%, p = 0.03) and cranial never injury (0.3 vs. 4.7%, p < 0.0001) was more common after CEA. ICSS resulted from the favorable results of CAVATAS and is also known as CAVATAS-2. ICSS⁸¹ randomized 1713 patients; in this study, the primary endpoint, similar to CREST, were more common after CAS (8.5% vs. 5.2%, p < 0.006). However, similar to SPACE and EVA-3S, there were issues regarding credentialing being more rigorous for the group performing CEA, as compared to CAS.

ICSS resulted from the favorable findings of CAVATAS and is also known as CAVATAS-2.⁸¹ It is a multinational prospective trial randomizing symptomatic patients equally suited for CAS or CEA. Additionally, lessons learned from EVA-3S are being applied. Attendance at a CAS training course is required, as well as mandatory proctoring for centers with limited experience admitted to the trial on a probationary status. Furthermore, embolic protection is recommended whenever the endovascular physician believes that a protection device can be safely deployed.

An additional ongoing study is the Asymptomatic Carotid Stenosis, Stenting versus Endarterectomy Trial (ACT I), a randomized trial consisting of low-risk patients with asymptomatic 80% to 99% carotid stenosis at multiple centers across North America.⁸² The primary outcomes will be 30-day MI, stroke, and death rates and 5-year stroke-free survival. Another study, the Transatlantic Asymptomatic Carotid Interventional Trial (TACIT) will randomize standard- and high-risk patients with asymptomatic carotid stenosis into one of three treatment arms: optimal medical therapy only (antiplatelet, antilipidemic, antihypertensive, strict diabetes control, and smoking cessation), optimal medical therapy plus CEA, or optimal medical therapy plus CAS with embolic protection.⁸³ Planned enrollment is 2400 patients with a primary end point of occurrence of stroke and death at 3 years. Secondary end points include rates of TIA and MI, economic cost, quality-of-life analysis, neurocognitive function, and carotid restenosis.

Most recently, preliminary results have been made available for the Parodi flow reversal system (Gore Neuroprotection System) in the Embolic Protection with Reverse Flow (EMPiRE) trial, which demonstrated a 30-day rate of TIA, stroke, MI, and death of 4.5%.⁸⁴ Additionally, the results of the Evaluating the Use of the FiberNet Embolic Protection System (EPS) in Carotid Artery Stenting (EPIC) trial demonstrated a 30-day rate of 3% for TIA, stroke, MI, and death (N = 237 patients) when using the FiberNet distal embolic protection system (Lumen Biomedical, Plymouth MN).⁸⁵ Continuing efforts and eventual completion of these trials or publication of the final results of these trials will improve our understanding of the relative indications for and contraindications to CAS and CEA.

Indications for Carotid Angioplasty and Stenting

The current indications for CAS are principally based on FDA and CMS approval for patients considered to be at high risk with CEA. As additional data become available from CREST, ICSS, ACT I, and other ongoing trials, further broadening of the indications for CAS is expected. However, at present, CAS should ideally be performed either in the setting of a patient with high surgical risk or after enrollment in an ongoing clinical trial. Guidelines for high surgical risk are principally derived from NASCET² and other CEA studies, as are indications for CAS. Therefore, CAS is approved for patients who are at high risk with CEA and have greater than 70% symptomatic carotid stenosis.

High-risk features for CEA include medical comorbid conditions, most prominently concurrent coronary artery disease causing angina and requiring CABG, recent or ongoing MI, and congestive heart failure (Table 351-4). Other medical risk factors extrapolated from NASCET include uncontrolled hypertension or diabetes and lung, liver, or renal failure. Additionally, CEA is considered high risk in patients older than 80 years. Surgical high-risk features include previous ipsilateral CEA or other perilesional surgery, previous neck irradiation, tracheostomy, or contralateral laryngeal palsy, all of which are related to the higher risk posed by surgical dissection during CEA. Anatomic considerations include a carotid bifurcation (lesion) located above the C2 vertebral body or below the clavicle or a very short neck and relatively high bifurcation, as well as severe neck arthritis causing a marked reduction in neck mobility during positioning for CEA. Another category of risk factors pertains to the available cerebrovascular reserve that renders CEA high risk, such as contralateral carotid occlusion and tandem lesions beyond the accessible cervical carotid artery. Similarly, neurological instability is also a recognized high-risk feature for CEA, including crescendo TIAs, stroke in evolution, TIAs while receiving heparin, multiple strokes, recent stroke, and acute occlusion.

TABLE 351-4 High-risk Features for Carotid Endarterectomy

Procedural Technique

CAS continues to evolve as a result of developing technology, evidence-based medicine, and operator experience. Here we describe our current routine for most CAS procedures; major steps for the procedure are illustrated in Figures 351-1 to 351-5. We briefly describe specific technologic considerations to entertain when planning CAS.

FIGURE 351-1 A 5 French diagnostic catheter is advanced into the external carotid artery on the affected side, and as the inset indicates, a stiff 0.35-inch exchange wire is left in the external carotid artery.

FIGURE 351-2 A 6 to 10 French guide is then brought into the common carotid artery over the exchange wire, and as the inset indicates, the wire and any introducer catheter are then removed from the guide sheath.

FIGURE 351-3 The lesion is then crossed with a 0.14-inch microwire (usually part of a distal embolic protection system). Once the lesion has been crossed, a distal embolic protection device is delivered through a microcatheter across the lesion and, as the inset indicates, unsheathed distal to the lesion.

FIGURE 351-4 After deployment of the distal protection device, a stent is brought up over the 0.14-inch microwire (left) and advanced across the lesion, where it is deployed, with the stenotic segment revealed as a waist apparent in the stent (right).

FIGURE 351-5 The stent delivery system is then retrieved, and an appropriately sized postdilation balloon is advanced over the wire and inflated briefly to its nominal pressure (left). Next, the balloon is retrieved and an angiogram is performed to confirm satisfactory revascularization. Once confirmed, the distal embolic protection device is recaptured and withdrawn (right).

The processes of angioplasty and stenting create intimal injury that promotes thrombosis.⁸⁶ Therefore, patient preparation with adequate antiplatelet and anticoagulation therapy is essential. Patients receive a dual antiplatelet regimen consisting of aspirin (325 mg daily) and a thienopyridine derivative (i.e., clopidogrel, 75 mg daily, or ticlopidine, 250 mg twice daily) for at least 3 days before stent treatment. A loading dose of clopidogrel (300 to 600 mg) administered early on the day of the procedure is an alternative for patients who are already taking aspirin. An intravenous bolus dose of heparin (50 to 60 U/kg) is administered after catheterization of the common carotid artery. An activated coagulation time of 250 to 300 seconds is maintained throughout the procedure. The heparin infusion is usually discontinued at the conclusion of the procedure.

The procedure is performed in an angiography suite with biplane digital subtraction and fluoroscopic imaging capability. The patient is sedated but arousable for neurological assessment. A Foley catheter and two peripheral intravenous lines are inserted. Blood pressure, oxygen saturation, and cardiac rhythm are monitored during the procedure. The carotid artery is generally approached percutaneously from the common femoral artery. The operator should also be familiar with radial and brachial approaches in the event that femoral artery access is not possible.

An aortic arch angiogram is obtained initially to define the atherosclerotic burden, as well as the anatomic configuration of the great vessels, which allows the operator to predict the feasibility of carotid cannulation and select the devices needed for the procedure. Selective carotid angiography is then performed and the severity of the stenosis defined. The diameters of the common and internal carotid arteries are measured, with attention paid to determining a landing zone for the embolic protection device. Intracranial angiography is also essential before the intervention because the presence of tandem lesions should be considered in the management strategy, as well as for comparison of preintervention and postintervention intracranial angiograms to confirm the absence of any vessel dropout suggestive of embolism.

Bradycardia occurs occasionally during angioplasty, and consideration may be given to administering glycopyrrolate (0.4 mg) before performing CAS. Atropine and vasopressors (e.g., dopamine and phenylephrine [Neo-Synephrine]) should be readily available should significant bradycardia and hypotension develop. Continuous intraprocedural monitoring of heart rate, blood pressure, and neurological status is essential. Therefore, we routinely transduce the arterial sheath in the femoral artery to maintain continuous arterial pressure monitoring.

After completion of the diagnostic angiogram and positioning of the catheter in the common carotid artery, road mapping of the cervical carotid artery is performed. An exchange-length 0.035-inch wire is positioned in the external carotid artery. The diagnostic catheter is exchanged over the wire for a 90-cm, 6 to 10 French sheath, which is then advanced into the common carotid artery below the bifurcation. For patients who have undergone complete diagnostic cerebral angiography before the stenting procedure, a combination of a 6 French, 90-cm shuttle over a 6.5 French Head-Hunter 125-cm slip catheter (Cook, Bloomington, IN) or a 5 French 125-cm Vitek catheter (Cook) can be used. In these cases, the shuttle is introduced primarily into the femoral artery over a 0.35-inch wire and parked in the descending aorta. The inner obturator and wire are removed. The 125-cm catheter is then advanced into the shuttle, and the target vessel is catheterized. The shuttle is brought over the wire and the catheter in the common carotid artery. The size of the shuttle is usually dictated by the profile of the embolic protection device and compatibility with the stent system. An optimal angiographic view that maximizes the opening of the bifurcation and facilitates crossing the stenosis should be sought. The lesion is crossed with the protection device. Predilation of the stenotic vessel segment is performed at the operator’s discretion. We avoid predilation whenever possible. If the lesion does require predilation, we prefer to undersize the balloon to simply facilitate crossing of the stent, usually with a 2- to 3-mm-diameter balloon. A 3- to 4-mm coaxial angioplasty balloon is advanced to the lesion over the 0.014-inch wire holding the protection device. On rare occasion, predilation needs to be performed before the introduction of an embolic protection device.

The diameter of the stent should be sized to the caliber of the largest segment of the carotid artery to be covered (generally 1 to 2 mm greater than the normal caliber of the common carotid artery). Oversizing of the stent in the internal carotid artery does not usually result in adverse events, but a tapered stent can better conform to the vessel wall. Particular attention should be paid to selection of a stent that is long enough to cover the entire lesion.

After removing the stent system, poststent dilation should be performed with a balloon that has a diameter matching that of the internal carotid artery distal to the stent. A coaxial balloon is generally preferred for this purpose. The embolic protection device is then removed via its retrieval catheter (when a balloon occlusion catheter is used for cerebral protection, the embolic debris is aspirated before deflation and retrieval of the balloon).

The femoral access site is then secured through either manual compression or a local clamping device; alternatively, for faster patient mobilization and greater comfort, a device may be used for closure (e.g., Mynx, AccessClosure, Mountain View, CA; StarClose, Abbott Vascular; Perclose, Abbott Vascular; or Angio-Seal, St. Jude Medical, Minnetonka, MN), the particular one being selected on the basis of operator preference, patient anatomy, and puncture site location.

Device Selection

A plethora of devices are currently available for CAS in the United States, and others are in the process of undergoing evaluation in FDA-approved trials. The current stent designs are centered around stent strut and tine sizes. The stents are constructed principally of annular rings interconnected by bridges. Tines refer to the size of each individual hole in the wall of the metallic alloy stent, which is determined by the number of bridges connecting the annular rings. The stents are either a nickel-titanium or cobalt-chrome alloy. They broadly fall into open-cell and closed-cell types. The smaller the stent tines and more connected the struts, the more the stent is a closed-cell design and the more rigid its overall structure. Therefore, closed cells tend to straighten the vessel after deployment and prevent the extrusion of plaque material through the tines because of an increased density of intima apposed to the stent wall. This feature makes them more appropriate for symptomatic lesions that probably have more unstable plaque with exposed debris. However, the rigidity of these stents reduces their conformability around the bends and angles frequently encountered in the cervical carotid artery, thereby further exaggerating the sharp bends and potentially worsening poststent kinking of the native vessel. Conversely, open-cell stent designs conform to the native vessels more effectively; however, the potential “cheese-grating effect,” particularly in symptomatic unstable plaque, may be a cause for concern. Of the closed-cell stents currently approved by the FDA, the Wallstent (Boston Scientific, Natick, MA) has the smallest cell size, followed by the Xact stent (Abbott). The NexStent (Boston Scientific), Acculink (Abbott), Protégé (ev3, Plymouth MN), and Precise (Cordis) are FDA-approved open-cell designs.

Distal protection devices started with the PercuSurge GuardWire occlusion balloon, which was simply a hypotube inflatable balloon on a 0.014-inch wire. At the completion of CAS, the guide catheter was used to aspirate the debris before deflating the PercuSurge balloon, which was placed distally in the internal carotid artery. Subsequently, several devices were developed that acted as filter devices with varying pore sizes to allow continued blood flow yet constrain the passage of embolic debris through the distal internal carotid artery where they were deployed. These devices can be considered for use in a particular patient on the basis of the following attributes: crossing profiles (i.e., the diameter when crossing the lesion), deformability and diameter to completely cover or seal the lumen of the internal carotid artery (thereby preventing flow around the filter), pore size and effectiveness in constraining embolic debris, and ease of deployment and retrieval, as well as loss of captured material during retrieval and induction of local carotid wall injury or spasm.

FDA-approved distal protection devices include the AngioGuard (Cordis), FilterWire EZ (Boston Scientific), AccuNet (Abbott), Emboshield (Abbott), Spider RX (ev3), and FiberNet (Lumen Biomedical) (Table 351-5). The AngioGuard has 100-µm pores and a crossing profile of 3.2 to 3.9 French. The FilterWire EZ has a crossing profile of 3.2 French with 80-µm pores. The AccuNet has 115-µm pores and a crossing profile of 3.5 to 3.7 French. The Emboshield has 140-µm pores and a 2.9 to 3.3 French crossing profile. Spider RX provides the additional option of being deployable across any 0.14-inch guidewire of the operator’s choice; however, its pore size varies, being smallest at the apex and larger at the periphery. All these devices are polyurethane membranes or nitinol mesh on a nitinol frame. The FiberNet has a very different design with radially arranged fibers that purportedly can filter particles as small as 40 µm without compromising flow. There are no clear data that support the use of one of these devices over another. We select the device on the basis of ease of use, crossing profile, and pore size.

The latest embolic protection system to receive approval from the FDA is the Parodi flow reversal system (Gore Neuroprotection System). This system consists of a distal balloon catheter placed in the external carotid artery and a balloon guide catheter placed in the common carotid artery below the lesion with a drainage mechanism that allows the reversed flow from the internal carotid artery to return to the systemic circulation through a filtered venous catheter. Another system currently being tested in the United States is the Mo.Ma proximal protection, flow reversal system (Invatec US, Bethlehem, PA) in the Proximal Protection with the Mo.Ma Device during Carotid Stenting (ARMOUR) trial. Both these systems obviate concerns of crossing the lesion without protection (including delivery of the distal protection device itself), as well as those associated with the true effectiveness of distal protection, as noted earlier. The concerns associated with proximal protection devices are their large size, currently requiring 9 French access to the femoral artery, and the concurrent delivery of a 9 French system into tortuous common carotid arteries.

Periprocedural Management

Hydration is essential before, during, and after the procedure. In patients who have concurrent renal dysfunction, hydration includes alkalinization of the urine with three ampules of sodium bicarbonate added to each liter of 5% dextrose in normal saline for prophylaxis against contrast-related nephropathy. In others, 5% dextrose in normal saline (0.9% NaCl) will suffice. Intravenous hydration should be maintained for up to 24 hours after the procedure. CAS at the carotid bulb often results in a sustained vagal bradycardia, which during the procedure may be minimized by the administration of anticholinergic agents such as glycopyrrolate. Postprocedure hypotension should be aggressively managed with vasoconstrictors such as Neo-Synephrine or intravenous infusion of dopamine. The goal is to keep the patient in a normotensive condition. Frequently, carotid artery disease is associated with coronary artery disease, and sustained hypotension is a harbinger of myocardial ischemia. Conversely, sustained hypertension after CAS revascularization may be associated with a cerebral hyperperfusion syndrome and, in some cases, with breakthrough hyperperfusion hemorrhage with disastrous consequences.⁸⁷ Hence, normotension should be continuously maintained with a systolic blood pressure between 110 and 150 mm Hg. This requires postoperative observation in a continuously monitored setting. If a closure device has not been used, the arterial sheath should be removed when the activated coagulation time is less than 150 seconds. The patient is usually discharged the afternoon after the procedure if maintaining hemodynamic parameters without intravenous infusions and if neurologically unchanged after the procedure. Patients require surveillance imaging to evaluate vessel patency. Duplex ultrasonographic evaluation should be obtained before discharge; at 6 weeks, 3 months, 6 months, and 1 year; and then annually thereafter. The dual antiplatelet regimen of aspirin and clopidogrel (or ticlopidine) is maintained for 12 weeks after the procedure, after which patients remain on aspirin therapy.

Procedural Durability

The durability of carotid revascularization with CAS is a concern frequently expressed by the surgical community. In a retrospective study of patients undergoing stenting for de novo (119 arteries) and postendarterectomy (76 arteries) carotid stenosis, 80% or greater stenosis was detected by follow-up Doppler imaging in 5.2% of the vessels stented.⁸⁸ Restenosis after endarterectomy was the major risk for in-stent restenosis. Significant (symptomatic or ≥80%) recurrent stenosis was detected by follow-up Doppler imaging in 6 (5%) of 112 patients in our CAS series.⁸⁹ The 3-year follow-up of the SAPPHIRE trial revealed a 4% recurrent stenosis rate after CAS.⁴⁹ This rate compares favorably with the 0% to 7.9% risk for restenosis reported after endarterectomy in large series.⁹⁰ Furthermore, an increasing body of evidence is indicating a persistent benefit after CAS. It appears that the majority of the risk associated with the procedure occurs within the periprocedural period and up to 30 days afterward. Thereafter, the risk falls significantly and remains low for the duration of these studies. SAPPHIRE suggested a 6.6% ipsilateral stroke rate at 3 years as compared with 5.4% after CEA, including the periprocedural risk (see Table 351-2).⁴⁹ The 2-year data from the SPACE trial suggest a 2.2% risk for ipsilateral stroke after CAS versus 1.9% after CEA (between postprocedure day 30 and 2 years).⁸⁸ Similarly, EVA-3S had a 1.26% stroke rate after CAS as opposed to 1.97% after CEA (between postprocedure day 30 and 4 years).⁵³

Complications and Their Management

Creative endovascular solutions can be found even for high-risk patients when treatment is deemed necessary. For patients with intraluminal thrombus and symptomatic carotid disease, the traditional treatment has been heparin and warfarin therapy with reevaluation in 6 weeks to 3 months. In four patients with multiple TIA episodes, we used a trapping technique with proximal and distal balloon occlusion with good success.^96,97 Flow reversal systems may prove useful in this setting and in the context of kinks that make landing of a DEP device infeasible.⁹⁸

Low-risk patients are those with either asymptomatic or single retinal or hemispheric TIAs and no previous cardiac history.²¹ Anatomically, type I aortic arches⁹⁹ with both straight proximal and distal anatomy provide the easiest anatomic substrate for CAS. Many of these patients have not been treated in the carotid stenting pool that has been reserved for “high-risk” patients. Ongoing low-risk clinical trials such as ACT I will determine how these patients fare with CAS versus CEA.

Complications Associated with the Steps and Tools for Carotid Angioplasty and Stenting

Distal Embolic Protection

The technique of DEP has lowered morbidity and mortality rates associated with CAS. Additionally, DEP devices provide excellent platforms on which to perform CAS. Nonetheless, each step of CAS, from crossing the stenosis to retrieval of the DEP device, has potential for complication. Transcranial Doppler (TCD) data have documented hits indicative of embolism at all stages of CAS; however, most TCD-detected emboli are clinically silent. On the basis of TCD data in which protected stenting with the PercuSurge device was compared with unprotected stenting, the highest risk maneuvers for emboli in unprotected stenting, in order from lowest to highest, were predilation angioplasty, stenting, and postdilation angioplasty.⁵³ In the PercuSurge-protected stenting group, guide sheath placement, guidewire manipulation, and deflation of the device were high embolic periods. DEP significantly lowered the risk of hits documented by TCD imaging. Stenting performed in conjunction with filter devices for DEP (FilterWire EX or AccuNet) in a series of 10 patients was monitored with a multifrequency TCD imaging system.¹⁰⁷ The system used was capable of distinguishing between gaseous and solid emboli. During deployment of the DEP device, more than 8000 microemboli were detected, with more than 40% being solid. During the stenting procedure, more than 7000 microemboli were detected, again with more than 40% being solid. In none of the patients did clinical sequelae develop. Microemboli occur even in the gentlest hands, but with good patient selection and technique, the occurrence of symptomatic embolic phenomena can be kept low.

The indication for proximal versus distal protection has not been determined. Logically, intraluminal thrombus, soft plaque, and a poor distal landing zone for DEP would be indications for proximal protection. The results of the European Imaging in Carotid Angioplasty and Risk of Stroke (ICAROS) study showed that gray-scale median (GSM) scores of 25 or less (representing echogenic plaque) are associated with higher embolic potential.¹⁰⁸ The ICAROS investigators created a prospective registry of 418 CAS cases from 11 centers and recorded GSM scores preprocedurally. Eleven of 155 (7.1%) patients with GSM scores of 25 or less had strokes versus 4 of 263 (1.5%) patients with GSM scores higher than 25 (P = .005). Taking this one step further, the ICAROS investigators validated the use of DEP in patients with GSM scores higher than 25 (P = .01), but not in those with GSM scores of 25 or lower. For these patients, stenting with proximal embolic protection devices or CEA may prove safer.

When crossing a carotid lesion, plaque dissection and embolization can occur. In the case of dissection, the patient is already therapeutically heparinized, and if the true lumen can be entered with the DEP device, it should be crossed and stented along with the stenosis. If the lesion cannot be crossed with the device, a microwire and catheter should be used to cross the lesion. The true lumen is identified with a microcatheter angiographic run, and the DEP device is then brought into position. Embolic phenomena associated with crossing the lesion require quick completion of the cervical carotid stenting portion of the procedure to gain access to the distal intracranial circulation. Inability to cross a lesion should lead to reevaluation of the procedure and consideration of the feasibility of endarterectomy.

Ideally, the filter should be deployed in a straight segment distal to the stenosis, well apposed to the carotid wall. Predeployment of the filter in the lesion can occur inadvertently. Most often this occurs secondary to the operator advancing the wire but not the housing sheath of the DEP device. When unsheathing a filter, if the filter wire is not fixed as the sheath is retracted, the entire system can be dragged through the lesion. In this case, the DEP device should be recaptured and redeployed distally. With monorail delivery systems, such maneuvers are not always possible. Pushing the filter back into the internal carotid artery above the lesion is frequently a better choice than dragging the entire system through a lesion. A secondary strategy for retrieval is to exchange out the initial DEP device sheath, run a 4 French angled catheter up the wire, and recapture the misdeployed device.

Loss of the filter device is rare, but possible. Filter devices have been sheared off with recapturing. In these dire settings, options include retrieval of the snare into the guide catheter, stenting the filter against the vessel lumen, or operative retrieval. If the device is in the bony segment of the skull base, operative retrieval may be impossible. Before any maneuvers are undertaken to retrieve a lost filter or other detritus, consideration should be given to arresting flow temporarily to avoid distal intracranial embolization.

In arteries in which a kink in the common carotid artery or internal carotid artery has been moved cranially with significant movement of the filter DEP device and in some particularly sensitive arteries, carotid vasospasm can occur. Generally, this is not a clinically or an angiographically significant problem, but stroke has been described in conjunction with sudden severe spasm just after deployment of an AngioGuard embolic protection device.¹⁰⁹ Spasm developed in three patients in one of the initial series in which the Emboshield was used, which was flow limiting in two patients.¹¹⁰ In non–flow-limiting spasm, the procedure should be completed and the filter recaptured. Usually, the spasm clears after the passage of a few minutes. Contrast material remaining within the arterial wall should cause concern for dissection. In flow-limiting situations or if significant time has passed and the spasm persists, intra-arterial nitroglycerin (100 to 200 µg) or verapamil (5 to 10 mg) is usually effective. Because these pharmacologic injections can certainly aggravate any ongoing hypotension, intravenous pressors should be at hand to maintain normotension. Distal dissection at the level of the filter can occur, more commonly so with the PercuSurge device than with filter devices. In patients with small, asymptomatic, and non–flow-limiting dissections, clinical observation is recommended. Stenting is warranted if the dissection is symptomatic or flow limiting. Spasm also occurs when a kink in the carotid artery is moved cranially by the DEP device and guide catheter. This can be ignored because it will resolve with retrieval of the device and often resolves after stenting and postdilation angioplasty.

Predilation angioplasty should be used only when necessary because it does carry with it some embolic risk from disruption of the plaque; such risk may be partly mitigated once the DEP device has been deployed. As mentioned, a 2.0 to 3.0 × 30-mm balloon is used. Attention should be paid to blood pressure reduction because of the carotid baroreceptor response. Patients who do not have pacemakers, are not undergoing CAS before CABG, and have initial baseline heart rates of less than 70 beats per minute may benefit from 0.75 mg of atropine or 0.4 mg of glycopyrrolate given at least 2 to 3 minutes before balloon inflation. Postdilation angioplasty is associated with a high risk for both embolic phenomena and severe baroreceptor response. Selection of balloon size is based on measurement of the stent in the distal internal carotid artery; the balloon is undersized by 0.5 mm. Angioplasty balloons in vessels in patients undergoing CAS before CABG are further underdilated to avoid a severe baroreceptor response. Dopamine or phenylephrine should be readied at the start of the procedure for rapid administration should the patient’s blood pressure drop significantly. It is important to remember that atropine may mitigate against severe bradycardia but will have little effect on hypotension.

Immediate complications associated with stenting are unusual. With dual antiplatelet therapy for 12 weeks and with arteries larger than 3 mm, acute and subacute thrombosis is uncommon. Subacute thrombosis has occurred twice at the University at Buffalo neurosurgery practice when cardiac surgeons stopped the patients’ antiplatelet medications within the first 2 weeks before CABG. Aspirin use was discontinued before urologic surgery in another patient 3 months after stenting of a carotid dissection with stents extending from the proximal segment of the internal carotid artery to the petrous segment of the internal carotid artery, and the stents became occluded; some degree of abnormal endothelialization was probably present in this patient, and a hypercoagulable state may have been implicated. A dual antiplatelet regimen for 12 weeks after the procedure and aspirin use for life appear to be essential. Drawing on the cardiology literature, early stent thrombosis is probably due to a dissection unrecognized at treatment or an undersized or expanded stent; late thrombosis is probably due to stent-artery mismatch, hypersensitivity, abnormal endothelialization, or poor compliance with antiplatelet medications.¹¹¹

If arterial occlusion occurs acutely during the procedure, the diagnosis includes severe spasm, dissection, thrombosis from plaque or platelet aggregation, and a filter that is filled with embolic material. As long as there is no dissection, the spasm will resolve. Nitrates or calcium channel blockers can be given, as mentioned earlier. Treatment is required for an occlusive dissection. A microcatheter and microwire need to be brought past the flap into the true lumen to accomplish this. If the clot has not embolized to the intracranial circulation and the patient is asymptomatic, consideration may be given to administering heparin to the patient overnight, checking collateralization with angiography, and performing stenting under proximal flow arrest. In an acutely symptomatic patient, stenting after the administration of a lytic agent and glycoprotein IIb/IIIa inhibitor may be necessary. If the filter is filled with embolic debris, a utility catheter can be used to perform a suction thrombectomy and the filter carefully captured and brought through the stent so that the captured debris is not disturbed. Successful outcomes have been reported in a small series of patients undergoing operative rescue after acute or subacute stent thrombosis.¹¹² However, if the experience with emergency CEA for treatment of acute carotid occlusion is a guide, the associated morbidity and mortality rate is greater than 20%.¹¹³ The hope is that with CAS, where we have the ability to visualize and access the entire cerebral vascular system, and with newly evolving techniques for acute stroke treatment, we will have better salvage procedures and success with endovascular approaches.

Recapturing the filter device is the final technical step of CAS with DEP. When this step goes smoothly, it is effortless; however, when accessing the filter with the recapturing sheath proves difficult, dissections, stent movement, and shearing of the filter device can occur. The most common setting for difficulty in recapturing the filter is with an open-celled stent on a significant curve or with the DEP device parked in a tortuous distal vessel. A systematic approach to recapturing maneuvers will generally lead to successful recapture: (1) advancing the guide catheter into the stent will bias the wire away from the stent wall, thereby allowing the recaptured sheath to pass; (2) having patients inhale deeply or turn their heads opposite the direction of the vessel curve can help straighten the curve or elongate the artery enough for passage of the sheath; more aggressively, pressing on the stent in the patient’s neck will also change the bias of the wire; (3) if the sheath is impeded by a stent tine, redilation with a larger balloon or spinning the sheath with forward pressure will help flatten the tine or allow passage of the sheath; and (4) if other maneuvers fail, a 4 or 5 French angled Glide catheter can be passed over the DEP device wire and used to capture the filter.

Intracranial Complications

The intracranial complications of carotid stenting can be grouped into large-vessel occlusion, shower of emboli, and hemorrhage. All three sources should be entertained in a patient with an acute or delayed neurological change not explained at the cervical level. In the acute setting, cerebral angiography should be performed to look for vessel cutoff or slow flow and emptying. If a clear large-vessel cutoff can be seen, an immediate attempt should be undertaken to recanalize the occluded vessel. A microcatheter (0.014-inch lumen or larger) should be brought through the lesion to confirm patency of the distal vessel and the length of the occlusion. In cases of acute stroke at the University at Buffalo, we have been using the microcatheter for the Merci device (Concentric Medical, Mountain View, CA) as the initial catheter to obviate the necessity for additional exchange maneuvers. One or two passes with the Merci retriever should be attempted. The other approved device in this setting is the Penumbra catheter system (Penumbra, Inc., Alameda, CA), which allows aspiration of distal debris and clot. If these devices are not available, thrombolytics can be used initially. There is some evidence that glycoprotein IIb/IIIa inhibitors give additive benefit.¹¹⁴ Other devices that can be used to open occlusions of large intracranial vessels (M1 segment of the middle cerebral artery, A1 segment of the anterior cerebral artery, P1 segment of the posterior cerebral artery, basilar artery, carotid artery) include balloon angioplasty, snares, large microcatheters, and stents.¹¹⁵

In settings other than clear large-vessel cutoff, if the DEP device has already been deployed, the procedure should be completed and a cranial CT scan obtained. Intracranial hemorrhage can be manifested as slow flow, and before expansion of the hematoma and development of a significant mass effect, it may appear as a shower of emboli. If an angiogram documents slow flow and the CT scan is negative for hemorrhage, glycoprotein IIb/IIIa antiplatelet agents are administered as a bolus dose followed by a 23-hour infusion. The patient’s blood pressure should be controlled tightly, with a systolic blood pressure of less than 160 mm Hg. If hemorrhage is identified, the systemic heparin anticoagulation should be reversed with protamine, blood pressure tightly controlled, and a repeat CT scan obtained in 6 to 12 hours. Life-threatening hematomas in neurologically salvageable patients can be evacuated. In the setting of hematoma expansion but no operative indication, activated factor VII can be given to stop progression of the hemorrhage. Most intracranial hematomas expand within the first 12 hours.¹¹⁶ Strong consideration should be given to stopping the patient’s dual antiplatelet therapy. The source of reperfusion hemorrhage remains debatable; some advocate a hyperperfusion origin, whereas others have suggested hemorrhagic conversion of a shower of emboli. In different cases, both sources are probably possible. However, once the hemorrhage has occurred, the treatment is identical.

Occasionally, a patient will have an intracranial aneurysm ipsilateral to a critically severe carotid stenosis. In a review of NASCET data, 1 in 90 patients with aneurysms known before the performance of CEA experienced aneurysm rupture (at 6 days after CEA).¹¹⁷ On the basis of the findings of the International Study of Unruptured Intracranial Aneurysms (ISUIA),¹¹⁸ the following approach seems rational: for patients with unruptured, asymptomatic aneurysms smaller than 6 mm, CAS can be performed before any treatment of the aneurysm, whereas for patients with unruptured, symptomatic aneurysms 6 mm or larger, the aneurysm should probably be treated before CAS is performed.

Systemic Complications

Systemic complications of endoluminal carotid intervention include seizures, MI, contrast material nephropathy, and contrast material allergy. If a patient has a seizure during angiography, attention should first be paid to basic life support and pharmacologic seizure control consisting of establishment of an airway and administration of lorazepam (2 mg) and a loading dose of phosphenytoin (18 to 20 mg/kg body weight). The differential diagnoses include embolism (air, necrotic debris), ischemia from vessel occlusion or vasospasm, intracranial hemorrhage, hyperperfusion, and contrast material sensitivity (which is less common with lower osmolar load agents). After the seizure has been aborted and control of the airway established, angiography should be performed, a CT scan obtained, and the differential diagnosis worked through until a source is identified. Electroencephalography is performed if necessary.

The rate of MI in the SAPPHIRE trial was 2.5% at 1 year, which was statistically lower than the risk for MI during CEA.¹² One of the major benefits of CAS in the major studies completed is the lower incidence of intraprocedural and postprocedural MI. However, MI still occurs in the CAS population. Standard measures, including nitrate infusion, beta blockade, and heparin administration, should be initiated if MI is diagnosed by cardiac enzyme studies and electrocardiography. A cardiologist should be consulted early because patients with Q-wave infarction seen on the electrocardiogram will usually require acute revascularization.

Acute allergy to contrast material is another setting in which airway safety should be ensured. The patient should be intubated if necessary. Methylprednisolone (120 mg given parenterally) and epinephrine (1 mg) are administered as necessary. Patients with known contrast media allergies are treated with 30 mg of prednisone at 12 hours and then 1 hour before the procedure, along with 50 mg of diphenhydramine (Benadryl).

The exact incidence of contrast-associated nephropathy is not clear. In one study of patients with creatinine levels lower than 1.5 mg/dL, only 8% had a 0.5-mg/dL increase in their level, and none had rises greater than 1.0 mg/dL with angiography.¹¹⁹ Other data suggest that contrast-induced nephropathy is the third most common cause of renal failure in hospitalized patients.¹²⁰ Certainly, the state of a patient’s renal function before angiography is the greatest determinant of post-treatment renal failure. A recent exhaustive review of contrast-related nephropathy¹²⁰ gave a few practical guidelines for renal prophylaxis in patients with preexisting elevated creatinine levels. Hydration with normal saline for 2 to 12 hours at a level of 1 mL/kg per hour before administration of contrast material is recommended. Low doses of low-osmolar contrast agents such as iodixanol should be given. Doses greater than 5 mL/kg of body weight divided by the serum creatinine level are associated with higher risk.

New Frontiers

Conclusion

CAS is a continually refined procedure that has much broader application potential than CEA does. With advancement in technology and experience reflected by improving outcomes in prospective randomized trials, one can expect the utility of this strategy to grow. However, for now, CAS is complementary to CEA and, for those of us who are trained in both techniques, provides a wide range of options to deal with a broad spectrum of pathologic conditions that affect the cervical carotid arteries.

Acknowledgment

We thank Paul H. Dressel, B.F.A., for assistance with preparation of the figures and Debra J. Zimmer, C.M.A.-A., for assistance with preparation of the manuscript.