Cerebral ischemia after carotid artery stenting

From the Center for Cardiology and Vascular Intervention, Hamburg, Germany

Carotid artery stenting has emerged as an alternative to endarterectomy for the treatment of carotid artery stenoses. Neuroprotective devices have been developed to help reduce the risk of stroke caused by procedure-related distal embolization. Diffusion-weighted magnetic resonance imaging (MRI) is a highly sensitive tool for the detection of cerebral ischemia. In unprotected carotid artery stenting preceded by three- or four-vessel cerebral angiography, silent cerebral ischemia was found in 37% of cases1 compared with only 25% of patients who underwent protected carotid artery stenting using the AngioGuard (J & J Cordis, Miami, Florida) filter device.2

In the present study, the incidence of cerebral ischemia was assessed using serial MRI in an unselected, consecutive patient cohort that underwent protected carotid artery stenting with a variety of devices without preceding diagnostic angiography.

Patients and methods

Of 50 consecutive patients who underwent carotid artery stenting at our institution, 42 consented to pre- and postinterventional MRI scanning of the brain. A total of 44 procedures were performed, corresponding to 44 hemispheres treated.

Protection devices. The cerebral protection systems used were the second- and third-generation NeuroShield (Abbott Vascular, Redwood City, California; n = 14), the FilterWire (Boston Scientific, Natick, Massachusetts; n = 14), the TRAP (Microvena Corporation, Plymouth, Minnesota; n = 11), the AngioGuard (n = 2), the Distal Protection Device (Medtronic AVE, Santa Rosa, California; n = 2), and the endovascular clamping device (MO.MA®, Invatec, Roncadelle, Italy; n = 1).

Carotid artery stenting procedure. At least 3 days prior to the intervention and until 4 weeks after the intervention, patients were treated with clopidogrel, 75 mg daily, and aspirin, 100 mg daily. Carotid artery stenting was performed with a 100-cm No. 5 French Vitek catheter (Cook, Bloomington, Indiana) and a No. 7 French long introducer sheath or the No. 11 French MO.MA system (in one patient). All patients were given a bolus of heparin, 70 to 100 IU/kg. The protection systems were placed, and the lesions were predilated. A self-expanding stent was deployed (predominantly using Wallstents, Boston Scientific), and the lesions were postdilated.

Magnetic resonance imaging. Within 24 hours before and after the intervention, MRI scans were taken in identical, single-shot, echoplanar sequences. The areas covered by hyperintensive foci were calculated, and patients with positive results were asked to undergo follow-up MRI scanning 3 to 6 months after the intervention. A calculation of the National Institutes of Health Stroke Scale was done by an independent neurologist as part of the neurologic examination before and after the intervention and at discharge.

Statistics. Comparisons of con-tinuous variables were made using the Mann-Whitney’s U test. P values

< .05 were considered statistically significant. Exact 95% confidence intervals (CIs) were calculated for proportions.

Results

In all patients, the protection devices were used as intended, and all procedures were completed successfully. One patient, in whom the TRAP device was used, developed a major stroke within 2 hours of the intervention; another patient, in whom the NeuroShield device was used, had a transient ischemic attack 3.5 hours after the intervention.

Before the intervention, MRI results were negative for all patients and showed hyperintensive foci after intervention for nine patients and 10 hemispheres (22.7%; 95%

CI, 11.5%—37.8%). Three of these patients had bilateral disease (23.1%; 95% CI, 5.0%–53.8%), and six patients had unilateral disease (20.7%; 95% CI, 7.8%–39.7%). Positive MRI results were shown for three of

13 patients with symptomatic lesions (23.1%; 95% CI, 5.0%—53.8%) and for seven of 31 patients with asymptomatic lesions (22.6%; 95% CI, 9.6%–41.1%).

Patients with positive MRI results did not have a higher percent diameter stenosis (91% ± 7% for positive results versus 86% ± 10% for negative results) or a significantly different crossing profile of the filter devices (1.23 ± 0.11 mm for positive results versus 1.32 ± 0.17 mm for negative results) compared with patients with negative MRI results. The procedure duration and the amount of contrast agent used were also not significantly different.

Eight of the nine patients who had positive MRI results did not experience any periprocedural neurologic complications. They had a small number of foci (median, 1; range, 1—3), with a mean size of 6.9 ¥ 2.7 mm2. The foci were located in the ipsilateral cerebral region in eight cases and in the contralateral circulation in one. In six of the nine cases with asymptomatic ischemic foci, follow-up MRI scans were obtained within a median of 4.2 months, with negative results in all cases.

One patient had a major stroke. In this patient, 12 hyperintense foci were found on the postinterventional MRI scan, with the largest focus (covering an area of 84.5 mm2) observed of any of the patients with positive results on MRI. All foci in this patient were located in the contralateral hemisphere, and five foci were seen on follow-up MRI scans

at 3.7 months, including the two largest foci. One patient experienced a transient ischemic attack (scotoma of the ipsilateral eye), which had no correlate on MRI scans.

Discussion

This study showed that focal ischemic lesions in the brain, as shown by diffusion-weighted MRI scanning, were found in 23% of patients who underwent neuroprotected carotid artery stenting. This in-cidence is similar to that reported in previous studies involving patients who underwent either cerebral angiography only3 or unprotected and protected carotid artery stenting preceded by cerebral angiography.1,2 In 90% of our patients, ischemic foci were not associated with periprocedural neurologic symptoms, and all asymptomatic lesions were reversible. Positive MRI results were independent of the presence of bilateral disease, symptomatic status of the patients, degree of the baseline stenosis, duration of the procedure, amount of contrast agent, and type of the protection device. Interestingly, focal ischemia was found, also in the contralateral cerebellum, in one of the asymptomatic patients in whom the MO.MA system was used.

One patient in our series experienced a major stroke. Postinterventional MRI scans showed a significantly higher number of ischemic foci, which were also larger compared with the asymptomatic patients. The largest of these foci persisted on follow-up MRI. The foci were also exclusively found in the contralateral hemisphere in this patient.

Our findings suggest that embolism was the underlying mechanism of cerebral ischemia. This is supported by analysis of the debris matter, which was collected or aspirated from carotid artery lesion sites.4,5 Several scenarios for embolism are conceivable. Before a protection system can be placed, guidewires, catheters, and sheaths must be manipulated within the aortic arch, which may result in the release of small particles of the often-calcified and atherosclerotic vessel wall. These particles would have had access to all supra-aortic vessels, which can result in ischemic lesions, including in the contralateral hemisphere, as was found in two of our patients.

Particulate matter may also have passed through the filter, either through the pores, which have a size of between 80 and 150 µm, or along the vessel wall in the event the filter did not have complete wall apposition. The final steps of the procedure, filter retrieval and withdrawal of the device, may also have released small pieces into the cerebral circulation.

Despite the use of cerebral protection devices, ischemic foci shown by diffusion-weighted MRI imaging were found in 23% of patients undergoing carotid artery stenting. Unavoidable interventional manipu-lation before the actual treatment of the target lesion may lead to serious neurologic complications. These findings show that interventionists should be extremely aware of adverse vessel anatomy and morphology. In addition, the medical device industry should concentrate not only on developing neuroprotective devices, but also on creating less injurious endoluminal equipment to be used for vessel access.

Conclusion

Symptomatic neurologic complications seem to be only the tip of the iceberg. Further studies showing the clinical significance of “silent” ischemia of the brain, which may be associated with an impairment of cognitive function as has been shown with patients after cardiac surgery, need to be performed.6