Carotid Artery Stenting: Do EPDs Level the Playing Field?

Carotid artery stenting (CAS) has variably fallen in and out of favor. Some have argued that CAS should be reserved for those patients who are at highest risk, not only for endartectomy (CEA), but if only treated medically. Unfortunately, none of the current trials or registries has been adequately designed to allow for equal comparison between the three treatment strategies — surgery, stenting, and medical therapy. Recent randomized trials comparing CAS and CEA contain significant methodological flaws and are thus inadequate for definitive comparisons, though they are thought provoking.

Perhaps the interventional community has focused on the wrong parameters to assess the efficacy of CAS. Rather than focusing on the high-risk “patient,” by initially focusing upon the lesion, rather than the patient, further stratification then becomes possible based upon the patient risk. It may be that general superiority of CAS is only realized in the low-risk lesion in the highrisk patient, though with the proper lesion/patient selection and in the right hands, CAS should yield equivalent or superior results. For other scenarios, CEA or medical therapy may yield lower adverse event rates.

Numerous studies have demonstrated that lesion characteristics: clinical, angiographic, and ultrasonographic can predict embolic risk. The risk of major adverse events associated with CAS in symptomatic patients is roughly between < 5–15%, though can be even higher in those > 80 years of age or those with other co-morbidities such as renal insufficiency. Furthermore, those plaques that are more echolucent are thought to contain a softer lipid core and are thus more likely to be symptomatic. This underscores the necessity to personally review all preangiogram images, such as duplex, magnetic resonance angiography (MRA), or computed tomography angiography (CTA).

Studies are conflicting on whether plaque morphology increases the risk of embolic complications with CAS. It has been thought that the use of embolic protection devices equalizes the periprocedural risk and negates the potential for embolization with complicated plaques. Although some studies have supported this concept, it is important to review embolization during CAS.

Transcranial Doppler (TCD) can be used to monitor embolic counts during CAS. Multiple studies have confirmed that every step of CAS involves embolization. The largest mean embolic counts occur during the stenting stage.1 Interestingly, TCD embolic counts have not correlated with clinical sequelae, such as transient ischemic attacks or stroke.

In an attempt to better characterize the debris that embolizes, Ohki et al examined ex vivo models of the carotid bifurcation. Using “casts” taken from CEA specimens, a model for CAS was created, (using a filter with a pore size of 120 microns), and the effluent was examined for particle number and size during simulated CAS. Embolic material correlated with plaque characteristics and stenosis severity. The number of particles ranged from 2–126, with a range of sizes from 120–2100 microns.² Similar experiments by Coggia et al demonstrated that embolism occurred at every step of the procedure with most emboli being < 60 microns.³

In order to minimize embolization into the intracranial vasculature, embolic protection devices (EPDs) have been designed. There are three basic types of EPDs: proximal occlusion, distal occlusion, and filter based. All can protect against macroembolization, but only the proximal and distal occlusion systems can protect against microembolization.

Distal embolic protection systems were commonly used in the early stages of CAS evolution. Concerns over vessel visualization and tolerance to cessation of antegrade flow have limited their use. However, TCD studies have shown that the microembolic count (MEC) is dramatically reduced with distal balloon occlusion. Similarly, with distal occlusion balloon deflation, there are a number of TCD “hits”; these likely represent small particles that could not be adequately aspirated from the empty space surrounding the balloon. Previously, vigorous flushing was performed to remove such debris; however, such flushing can embolize the contralateral carotid or either vertebral artery.

Currently, filter-based EPD are most commonly utilized. Pore size varies between different filters, ranging from 100 to just above 200 microns. Designs vary, anddelivery systems and the degree of eccentricity differ. Each manufacturer claims to have superiority in filter design. There have been no prospective, randomized in vivo studies comparing devices. Ex vivo studies have been done, testing varying degrees of tortuosity. Transcranial Doppler MEC are thought to be higher with filter-based systems and this likely reflects the inability to capture debris smaller than the pore size.⁴

Proximal occlusion systems rely on the ability to reverse flow in the carotid artery system. These systems have the potential to provide the best protection for the high-risk lesion. Unfortunately, they are not widely utilized and require larger delivery systems.

Most importantly, do EPDs prevent strokes during CAS? There have been no randomized trials comparing CAS, with and without EPD and there likely never will be such a trial. There have been interesting studies, however, evaluating diffusion weighted MRI (diffusion-weighted imaging [DWI]) after CAS with and without EPD. In one study, 30% of patients developed a new focal lesion on DWI: 26% of patients with an EPD and 36% of patients without an EPD.⁵ In an attempt to delineate when these events occur, another study performed DWI pre CAS, immediately post CAS, and 48 hrs post CAS.⁶ Of the case, 9% had a new lesion immediately post CAS, but 67% had new lesions at 48 hrs. Interestingly, of the patients with new lesions at 48 hrs, 97% had ipsilateral lesions, 28% also had bilateral lesions, and there was one case of only a new contralateral lesion. The Global Carotid Artery Stent Registry found a stroke rate of 6.15% in patients without an EPD and 2.85% in patients with an EPD. However, the Pro-CAS registry found no difference in over 3000 CAS procedures. The ARCHeR trial also found no reduction in embolic events when an EPD was utilized. ARCHeR 1 did not require any EPD and had a lower stroke rate than ARCHeR 2 or 3, (though not statistically significant).⁷

Perhaps the most thought-provoking data comes from the timing of embolic complications of CAS. Analyzing the CAPTURE registry shows that 23% of strokes are non-ipsilateral (ipsilateral, 3.9% vs. contralateral, 0.9%). Furthermore, only 23% of strokes occur during the procedure, 58% occur after the procedure, and 20% after the patient has been discharged. Of the post CAS strokes, 38% occur > 24 hours after the procedure.

The role of CAS continues to evolve. Both the clinical understanding as well as the basic science behind CAS continues to advance. Currently, there are more questions raised than answered regarding CAS with EPD. All steps of CAS are associated with embolization, and although TCD can detect such, the TCD MEC has not correlated with cerebrovascular complications. Furthermore, a better understanding is necessary to explain the fact that only a minority of post CAS neurological events occurs during the procedure, whereas 78% occur after the procedure.