Transcervical carotid stenting with flow reversal protection: Experience in high-risk patients

Patients and methods 

From January 2005 to June 2006, we performed 62 CAS procedures in cases of carotid stenosis >70%, using transcervical access and carotid flow reversal for cerebral protection. Patients in whom carotid revascularization was indicated and who were at high risk for carotid endarterectomy (CEA) according to the Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE)6 criteria were prospectively and consecutively included in the study. All patients gave informed, written consent for participation. Results were retrospectively analyzed to determine outcomes.

The grade of carotid stenosis, presence of extensive calcifications, and morphometry of the lesion were determined by Doppler ultrasound on a Philips HD 11 unit (Bothell, Wash). Magnetic resonance angiography (MRA) was used to assess the brain parenchyma, morphology of the lesion, presence of vessel tortuousities, and the status of intracerebral circulation. All patients underwent a preoperative TCD study (PMD-100, Spencer Technologies, Seattle, Wash) to evaluate the cerebral vascular reactivity (considered exhausted when the mean velocity in the middle cerebral artery increases <20% after 20 seconds of apnea),7 the presence of high-intensity transient signals (HITS), and the existence of associated intracranial lesions. All patients underwent a complete neurologic examination before and after the procedure by a stroke neurologist, who was also present during all the procedures.

In the last 35 patients in the series, TCD monitoring was performed during revascularization to determine the presence of HITS and variations in the middle cerebral artery flow velocity and pulsatility index at the start and completion of the procedure, immediately after clamping of the common carotid artery and a few minutes after clamping. This was done to analyze the mechanisms of cerebral compensation. These data are presented in a separate article8 that mainly analyzes intracerebral hemodynamic aspects but not technical issues and clinical outcome, which are the subject of the present study.

All patients were prescribed acetylsalicylic acid (300 mg/d) and clopidogrel (75 mg/d) for at least 4 days before the procedure. In cases of noncompliance with this treatment, a 300-mg loading dose of clopidogrel was administered 24 hours before the procedure.

The surgical technique described by Criado et al9 was used in all patients. Under local anesthesia, a vertical mini-incision was made in the base of the neck to access the proximal common carotid artery, which was controlled by a vessel loop. After dissection of the internal jugular vein, a shunt was created between the carotid and jugular vein by placing an 8F introducer sheath (Super Arrow Flex; Arrow International, Reading, Pa) in each vessel and connecting the introducers with a 15-cm tube (Fig 1, Fig 2). Under systemic heparinization, the common carotid artery was occluded, and retrograde flow was established in the internal carotid artery with fluoroscopic control to ensure proper function of the shunt. TCD monitoring confirmed correct flow inversion in the intracranial portion of the internal carotid artery in patients in whom it was possible to examine this segment of the vessel; HITS were not seen in any of these cases. Angiography in anteroposterior and oblique views was performed to quantify the stenosis.

View Large Image Download to PowerPoint

Fig 1. Under local anesthesia, a vertical mini-incision was made in the base of the neck to access the proximal common carotid and the internal jugular vein. A shunt was created between the carotid and jugular vein by placing an 8F introducer sheath in each vessel and connecting the introducers with a short tube.

View Large Image Download to PowerPoint

Fig 2. A shunt was created between the carotid artery (1) and jugular vein (2) by placing an 8F introducer sheath (3) in each vessel and connecting the introducers with a tube. The common carotid artery was occluded (4), and retrograde flow was established.

Next, a 0.014-inch (190 cm) Spartacore (Guidant, Santa Clara, Calif) or 0.014-inch (180 cm) Platinum Plus ST guidewire (Boston Scientific, Natick, Mass) was inserted in a 40-cm long Berenstein 4F catheter (Angiodynamics, Queensbury, NY), and the internal carotid artery lesion was crossed through the sheath placed in the common carotid artery. Once the proximal guidewire reached the carotid siphon, the stent (Acculink, Guidant; or Carotid Wallstent Monorail, Boston Scientific) was deployed and dilated with a 5-mm × 30-mm or 6-mm × 30-mm Viatrac Plus (Guidant) or a 5-mm ×20-mm or 6-mm × 20-mm Ultrasoft angioplasty balloon (Boston Scientific).

Slow, manual aspiration (20 mL) was done before the clamp was released, and an angiographic study was done to assess the technical outcome and investigate possible complications. Atropine (0.5 to 1 mg) was injected in the case of important bradycardia, and intra-arterial nitroglycerin (100 to 200 μg) was used when spasm of the distal carotid occurred. After withdrawal of the introducers, the puncture sites were closed with 5-0 polypropylene sutures.

Double antiplatelet therapy (aspirin and clopidogrel) was maintained for the first 30 days; thereafter, clopidogrel (75 mg/d) was prescribed indefinitely. Clinical and Doppler ultrasound follow-up studies were performed at 24 hours, 1, 3, 6, and 12 months, and yearly thereafter.

Results 

The mean age of the participating patients was 76.5 years (range, 56 to 92 years) and 52 (83.9%) were men. The distribution of cardiovascular risk factors and associated diseases is summarized in Table I. The main indications for carotid stenting were elevated cardiac risk in 28 patients (45.1%), of whom 4 had prior myocardial revascularization; severe obstructive pulmonary disease in 4 (6.4%), contralateral laryngeal nerve paralysis in 4 (6.4%), recurrent stenosis after CEA in 2 (3.2%), prior cervical radiotherapy in 2 (3.2%), distal internal carotid lesion in 1 (1.6%), and age >80 years in 21 (33.8%). None of the patients treated had creatinine values >1.5 mg/dL. Twenty-four patients (38.7%) presented with symptomatic >70 % carotid stenosis, of whom three had amaurosis fugax, 11 had hemispheric transient ischemic attack (TIA), and 10 had previous ipsilateral stroke.

Table I. Epidemiologic and clinical variables of the patients treated

		N	%
	Age, mean years (range)	76.5(56-92)
	Sex (male)	52	83.9
	Hypertension	50	80.6
	Diabetes mellitus	21	33.9
	Hypercholesterolemia	34	54.8
	Smoking	12	19.4
	Ischemic heart disease	38	61.3
	Pulmonary disease	19	30.6
	Renal failure	0	0
	Symptomatic	24	38.7
	Contralateral stenosis >50%	20	32.7
	Intracranial lesion	24	38.7
	Decreased or exhausted hemodynamic reserve	12	19.3

No stenosis was present in the contralateral carotid artery in 16 patients (25.8%), and in the remaining patients, stenosis was 30% to 50% in 16 (25.8%), 50% to 70% in seven (11.3%), and >70% in six (9.6%). Seven (11.3%) had occlusion of the contralateral internal carotid artery. The contralateral carotid had been previously revascularized in 10 patients (16.1%). In the TCD study before carotid stenting, an intracranial artery lesion was detected in 24 patients (38.7%), HITS in one (1.6%), and a decreased or exhausted cerebral vascular reactivity in 12 (19.3%). All patients except one received the appropriate antiplatelet therapy before revascularization.

The mean duration of the procedure was 50 minutes (range, 30 to 60 minutes), and the mean flow reversal time was 15.1 minutes (range 5 to 36 minutes). The mean volume of contrast used was 36 mL (range, 10 to 100 mL), and mean duration of fluoroscopy was 7.4 minutes (range, 3 to 19 minutes). The arteriographic control showed flow reversal in all cases. Balloon occlusion of the external carotid artery was not performed in any patient. Predilation of the lesion with 3-mm to 4-mm × 2-cm balloons (Viatrac Plus, Guidant) was performed in five patients (8.3%).

The self-expanding stents used were 6-mm × 8-mm × 30-mm, 6-mm × 8-mm × 40-mm, 7-mm × 10-mm × 30-mm, or 7-mm × 10-mm × 40-mm Acculink (Guidant) in 33 cases (54%), and 28% received an 8-mm × 29-mm, 9-mm × 40-mm, 10-mm × 24-mm or 10-mm × 37-mm Carotid Wallstent (Boston Scientific). Postangioplasty dilation was done in all cases.

Significant distal internal carotid spasm was observed in two patients, and was resolved successfully with intra-arterial nitroglycerin. Transient bradycardia in response to postangioplasty dilation was observed in nine cases (14.7%). One patient presented persistent hypotension during the first 12 hours and required noradrenaline infusion.

The procedure was successfully completed in 60 of the 62 patients, yielding a technical success rate of 96.8%. In one patient, the procedure could not be performed because it was impossible to cross the very extensive, preocclusive internal carotid lesion. The patient’s clinical condition was such that the risk of surgery was acceptable; hence, CEA was done. In the other patient, a major dissection was seen in the common carotid artery after the stent placement was completed. Given its proximity to the introducer sheath, it did not seem advisable to attempt endovascular treatment, and a bypass from the common carotid artery to the distal internal carotid was performed instead. Neither of these two patients presented neurologic symptoms at the end of these procedures. None of the remaining patients had residual stenosis >30%.

One patient experienced a TIA of the anterior cerebral artery territory that recovered ≤12 hours. Another patient experienced a stroke with contralateral hemiplegia. A Doppler ultrasound examination 24 hours after CAS showed an in-stent thrombosis, and a cerebral computed tomography (CT) scan showed acute infarction signs in the middle cerebral artery territory ipsilateral to the revascularized carotid. In agreement with the stroke neurologist, we did not believe a new attempt of revascularizing this territory was indicated because of the recent infarction. At 48 hours after hospital discharge, another patient returned with hemiplegia and aphasia after an episode of intense headache. A CT scan showed an extensive cerebral hematoma that required surgical drainage. Thus, neurologic morbidity was 4.9% in this series.

Intolerance to flow reversal was observed in one patient (1.6%), who lost consciousness a few minutes the common carotid was clamped. The procedure was extremely rapid in this case, and the patient recovered without sequelae after declamping. One large cervical hematoma was observed, which required surgical drainage (Table II). TCD monitoring during the procedure detected HITS in two of the studied patients (5.7%) and flow reversal in the anterior cerebral artery in 28 patients (80%); no patient showed middle cerebral artery flow reversal. A significant improvement in the middle cerebral artery mean flow velocity and pulsatility index was observed at completion of the procedure.

Table II. Technical complications, morbidity, and neurologic complications related with transcervical access

		N	%
	Technical success	60/62	96.8
	Neurologic morbidity	3/62	4.9
	Complications
	Major (cervical hematoma)	1/62	1.6
	Minor	0	0

The 24-hour Doppler ultrasound examination confirmed stent patency in all patients except for the one mentioned patient (98.4%). In addition, during the first 24 hours, troponin levels were determined every 6 hours in all patients, and all results were negative.

Patients were discharged from the hospital 48 hours after the intervention. Median follow-up was 6 months (range, 1 to 18 months). No deaths or cardiac complications occurred in the first 30 days. There were two cases of significant restenosis (50% to 70%) and one restenosis of >70%. None of the patients presented neurologic events in the stented carotid vascular territory during follow-up.

Discussion 

During the last decade, there has been a considerable rise in the use of endovascular techniques to treat carotid stenosis and increasing interest in the development and optimization of the systems used to perform these procedures. Initial results with this approach were discouraging because the morbidity and mortality rates were not comparable with those obtained with CEA. Nevertheless, the introduction of routine stenting and cerebral protection systems considerably improved the outcome,10 and current results are similar to those reported for CEA.6, 11, 12

No comparative studies are available on the optimal cerebral protection system for CAS, and the published results from series using different techniques are similar. However, studies with TCD have shown that when distal protection systems are used,13 HITS are detected in most steps of the procedure. Although the clinical relevance of these findings remains to be determined, the association between HITS and the incidence of intraoperative stroke seems clear. Indeed, a study by Schüter et al14 of CAS procedures with protection devices found diffusion lesions indicating ischemia in up to 22.7% of patients when studying the cerebral parenchyma with diffusion-weighted magnetic resonance imaging.

In contrast, systems based on internal carotid artery flow reversal present several advantages compared with the use of distal protection systems. The Parodi Antiemboli System (PAES; ArteriA, San Francisco, Calif) transfemoral flow inversion catheter is the only one that has shown virtual elimination of the embolization particles in studies performed in vitro15 and in vivo.16 In our experience with the first 23 TCD-monitored patients,8 no air or solid emboli were detected at any phase of the procedure, confirming the hypothesis that internal carotid artery flow inversion offers adequate cerebral protection in most cases. In the remaining 12 patients monitored with TCD, HITS were observed in two patients. We believe that monitoring the shunt by fluoroscopy is fundamental for confirming reversal of flow, but TCD monitoring further allows detection of cases in which the flow reversal does not suffice to completely eliminate microemboli, resulting in incomplete cerebral protection.

Without external carotid occlusion, reliable reversed flow cannot be guaranteed throughout the procedure because internal carotid stump/back pressure can be quite low and venous pressure quite high with respiratory variation and coughing, causing intermittent antegrade internal carotid artery flow from the higher pressure external carotid system. Hence, fluoroscopy confirmation of reversed flow with a contrast injection does not confirm total reverse flow during the whole procedure. TCD is used to indicate when manual aspiration is necessary.

In our experience, the patients in whom the external carotid artery could not be visualized on diagnostic angiography once the common carotid was clamped had a greater probability of presenting HITS. In the two mentioned patients with HITS, manual transient aspiration was used at the time the lesion was crossed, at stent deployment, and at postangioplasty dilation. The disappearance of microemboli in the intracranial internal carotid on TCD demonstrated the efficacy of this maneuver. Therefore, we consider that transient aspiration, which simplifies the procedure and decreases possible complications, could be an alternative to balloon external carotid occlusion, as done by some authors.17, 18, 19 Recently, Alexandrescu et al20 described a variation of the technique that combines proximal occlusion of the common carotid and distal filter placement with transient aspiration applied during navigation over the target lesion and before filter deployment. The authors concluded that this maneuver might decrease the neuroembolic risk of the procedure.

Among the most controversial issues of proximal carotid occlusion with flow reversal is the patient’s tolerance of this technique. In the Criado series,5 clamping of the common carotid artery was well tolerated in 96% of patients, and only two of the seven patients with contralateral carotid occlusion did not tolerate clamping. Intolerance was seen in 3% in the Parodi et al4 study, and all the patients studied tolerated clamping in the Chang.18 In our experience, neurologic intolerance during the procedure occurred in only one patient (1.6%), who had contralateral carotid occlusion and obliteration of the anterior cerebral artery.

Preoperative evaluation to determine the status of intracranial circulation and the cerebral vascular reactivity by magnetic resonance angiography and TCD can be an aid to identify patients who may not tolerate carotid clamping. We found that patients with an exhausted cerebral vascular reactivity before the intervention presented poor collateral flow recruitment, and the middle cerebral artery flow velocity did not gradually increase as a compensatory mechanism after common carotid clamping.8 In our opinion, the two possible options to follow when this occurs are to remove the common carotid artery clamp and place a distal filter device, or (when possible) reduce the duration of flow inversion by performing the procedure quickly.

Neurologic morbidity in the present series was 4.9% (1 TIA, 1 ischemic stroke in the immediate postoperative period, and 1 cerebral hematoma), a rate similar to reported values in other published series.4, 5, 19 No deaths occurred in the first 30 days, and even though 45.1% of the patients presented with severe ischemic heart disease, no myocardial infarctions occurred among the treated patients. In addition, given that these were patients at high risk for CEA and thus excluded from multicenter reference studies, we believe the neurologic morbidity rate is comparable with that of conventional surgery.

The main advantages of transcervical CAS are summarized in Table III. One of the greatest advantages of transcervical access compared with transfemoral access, whether for flow inversion or for placement of a distal protection device, is that the aortic arch is not instrumented. Performance of an aortic arch arteriogram, cannulation of the brachiocephalic trunk or common carotid, and insertion of introducer sheaths or catheter guidewires in the common carotid are necessary steps in the femoral approach, during which the brain has no protection. These steps are not required with the transcervical approach.

Table III. Advantages of transcervical carotid angioplasty and stenting

	1.Target lesion is not crossed without protection 2.No instrumentation of aortic arch •Aortic arch types II and III •Bovine trunk •Tortuous supra-aortic vessels 3.Avoids difficulties of femoral access 4.Shorter duration of radiation exposure 5.Smaller volume of contrast 6.Shorter duration of procedure 7.Lower cost

Another point to be mentioned is the technical failure rate of up to 5% with the transfemoral route because of vessel tortuousities and type II or type III aortic arches in which the carotid arteries cannot be cannulated or reached because of associated aortoiliac pathology.21 In addition, the volume of contrast and exposure time reported here are much lower than the volumes required for diagnostic brain angiographies reported by other authors.22

Finally, the potential for puncture site complications, which can reach 15% for femoral access, is avoided.5 Cervical hematoma is probably the most frequent local complication associated with the transcervical approach, because this area is well perfused and patients undergoing the procedure are receiving antiplatelet therapy. Nevertheless, direct suture of the incisions made for the introducer sheaths minimizes the possibility of this complication developing. In the present study, drainage of a cervical hematoma was required in only one case in which the procedure was done on the left side. There was some difficulty in cannulating the internal jugular vein in this patient, a problem we have encountered with some frequency when working in the left side and which we attribute to the anatomic positioning of the large veins in this region.

In one patient severe dissection of the common carotid artery was observed at the end of the procedure, which required a surgical bypass. This patient had severe calcification of the proximal common carotid artery; hence, in agreement with Criado et al5 and Chang et al,18 we believe this should be one of the contraindications for the use of this technique.

Technical success was 96.8% in the present study. The incidence of technical failures in endovascular procedures for the internal carotid artery varies from 0%5, 19, 20 to 10%.23 According to Chang et al,18 up to 6% of these are due to the transfemoral route and, particularly, to the inability to insert the catheter and guidewire in the common carotid artery. The transcervical approach simplifies the procedure, and because access to the target lesion is shorter, the possibility of successfully completing the procedure is thus greater.

Conclusion 

Randomized controlled trials are needed to determine which cerebral protection system provides optimal performance during carotid stenting. In the interim, results from clinical series can serve as a guide to improve clinical practice. In our experience, flow reversal using a transcervical access route is a simple, safe method that eliminates the drawbacks of aortic arch instrumentation and crossing the target lesion without protection. The short-term and long-term outcomes are also good and appear comparable with the reported results for carotid revascularization by endarterectomy.

Manuel Quintana from the Section of Neurology of Hospital Universitario Vall d’ Hebron was responsible for the statistical analysis.

Author contributions