Transcervical carotid stenting with carotid artery flow reversal: 3-year follow-up of 103 stents

Patients and methods 

From March 2003 to July 2005, 103 consecutive transcervical CAS procedures were performed in 97 patients, consisting of 52 stents on the right carotid and 51 on the left. The first 50 cases included in this study were the matter of a previous report that documented the initial feasibility and safety of the procedure.12 Medical history, comorbidities, neurologic status, degree of carotid stenosis, intraoperative findings and events, and postoperative follow-up were prospectively recorded. Patient mean age was 72 years (range, 54-90 years), and 82 patients (79.6%) were men. Comorbid medical conditions are summarized in Table I.

Table I. Medical comorbidities in 97 patients undergoing 103 transcervical carotid stenting procedures

	Medical Comorbidity	No.	%
	Hypertension	81	79
	Hypercholesterolemia	58	56
	Coronary artery disease	35	34
	Smoking	25	24
	Diabetes mellitus	22	21
	Severe COPD	8	8
	Renal insufficiency	11	11
	Cigarette smoking	25	24

COPD, Chronic obstructive pulmonary disease.

Thirty-seven patients (36%) presented with ipsilateral (<6 months old) stroke (14%) or transient ischemic attack (TIA; 22%). In addition, 43 patients (42%) had a history of stroke (22%) or TIA (20%) more than 6 months before the procedure. Preoperative mean ipsilateral internal carotid peak systolic velocity (PSV) was 314 cm/s (range, 0 to 904 cm/s), and the mean contralateral internal carotid artery PSV was 124 cm/s. One patient with 0 velocity was suspected to have an open internal carotid, but a velocity waveform could not be obtained because of plaque calcification. A very tight lesion was confirmed during intraoperative angiography. The duplex estimated degree of stenosis was >50% for symptomatic patients and >70% for asymptomatic patients.

Patients were offered the procedure because of the higher risk for general anesthesia or of a major procedure and on their preference for a less invasive approach. All cases were done on the basis of preoperative duplex ultrasound imaging without preoperative angiography.

All procedures were done exclusively by vascular surgeons. Local/regional anesthesia was used in 72 cases (70%) and general in 31 (30%). Most of the general anesthetics (n = 25) were used during the first half of this experience while the technique was developed. Mean operative time was 69 minutes (range, 20 to 180 minutes). Mean administered contrast agent volume was 51 mL (range, 10 to 120 mL). All patients were taking clopidogrel before stent placement. Systemic heparinization was done with 100 IU/kg of body weight of intravenous heparin.

Our technique for transcervical CAS and interventional protocol have been previously described.12, 14 A mini-cervical cutdown was used to establish a carotid artery-to-jugular vein fistula by using commercially available vascular access sheaths (Fig 1).

View Large Image Download to PowerPoint

Fig 1. Diagrammatic representation of the technique used for transcervical carotid flow reversal during carotid stenting. (Illustration by Juan Fontcuberta, MD.)

Intraoperative diagnostic digital arteriography using hand injection was performed in the lateral and oblique planes to localize and quantitate the degree of ICA stenosis in all cases, revealing a 50% stenosis in six (6%), >70% in 75 (73%), and subocclusive stenosis in 22 (21%). With carotid artery flow reversal in place, a 0.014-in guidewire (Platinum plus TM, ST, 0.014-180 cm; Boston Scientific, Meditech, Miami, Fla) in a 4F or 5F × 40-cm-long Bernstein catheter (Angiodynamics, Queensbury, NY) was used to cross the ICA stenosis under fluoroscopic guidance, and the tip of the guidewire was advanced to the level of the carotid siphon. Predilatation of the carotid stenosis with 3- to 5-mm-diameter balloons was done in 34 cases (33%). Predilatation was done at the discretion of the operator based on the estimation that the stent delivery system would not cross the lesion without predilatation.

Self-expandable 7-, 8-, 9- or 10-mm diameter × 25-, 30-, or 40-mm-long stents were used, totaling 99 biliary Wallstent, Monorail stents (Boston Scientific, Maple Grove, Minn), and five Exponent (Medtronic, Minneapolis, Minn) stents. Two stents were deployed in one carotid artery. Poststent dilation was performed for 5 to 15 seconds in all cases with 5-, 5.5-, or 6-mm × 2-cm-long monorail balloon catheters (Ultra-soft SV Monorail Balloon catheter; Boston Scientific) inflated to 8 atm. Completion arteriography was done in all cases to assess technical results and the presence of distal spasm. Intra-arterial papaverine solution (1 mg/mL) was selectively used to treat residual carotid spasm.

After completion, the access sheaths were removed, the vessel access sites were closed with 5-0 or 6-0 polypropylene suture, and the cervical wound was closed with absorbable suture. All patients remained under observation in the recovery room with electrocardiographic monitoring after the intervention and were transferred to a floor or telemetry bed when stable. Troponin levels were not routinely obtained postoperatively, but were not elevated in two patients with transient asystole in response to balloon dilatation.

Clopidogrel was continued at 75 mg/d orally for at least 1 month, and aspirin was continued indefinitely.

Postoperative neurologic examinations were performed by vascular surgeons, residents, fellows, and recovery room nurses. A postoperative physical examination and a carotid duplex scan were repeated in all patients at 1, 6, and 12 months, and yearly thereafter.

Technical failure was defined as inability to access or to cross the lesion or inability to complete the procedure for any reason. A TIA was defined as a focal hemispheric deficit that resolved ≤24 hours, and a focal deficit lasting >24 hours was defined as a stroke. Recurrent or residual in-stent stenosis was defined as >50% diameter reduction as determined by duplex ultrasound scanning. Follow-up was conducted at 1, 6, 12, 24, and 36 months after surgery and included bilateral carotid artery duplex ultrasound imaging.

Results 

No patients died. Technical success was achieved in 100 (97%) cases. The three technical failures (3%) were immediately converted to carotid endarterectomy under general anesthesia: one because of a common carotid dissection with the entry sheath, one because of the inability to cross a very tight and angulated ICA origin lesion with the guide-wire, and one because of severe patient agitation that required conversion to general anesthesia and the surgeon chose to proceed with an endarterectomy rather than to pursue the stenting procedure.

There were no major strokes. Neurologic complications occurred in four patients and included one ipsilateral motor TIA, one contralateral TIA, and two minor strokes (2%). One of these was in a patient with a previous ipsilateral stroke who developed worsening hemiparesis, and another patient sustained dysarthria. Both patients returned to their baseline neurologic status within a week. Two neurologic events were apparent in the operating room and two in the recovery room. No in-hospital myocardial infarctions were recorded; however, routine troponin level determinations were not obtained in the patients. The major adverse eventrate was 0%.

Completion arteriography revealed one ICA (30%) residual stenosis and demonstrated that the external carotid artery remained patent in all cases. Among the first 12 cases, four common carotid artery dissections occurred at the access site. One required surgical repair with a proximal common carotid interposition graft, and three resolved after placement of the stent intended to treat the stenosis without additional stenting. During the remaining 91 procedures, one common carotid artery dissection occurred at vessel entry, which prompted one of the conversions to endarterectomy. Two postoperative wound hematomas required surgical drainage under local anesthesia, but neither required blood transfusion nor delayed the patient’s discharge from the hospital. No wound infections occurred.

Mean carotid flow reversal time was 21 minutes (range, 7 to 60 minutes). Three of the 72 patients (4%) who underwent the procedure with local anesthesia did not tolerate carotid flow reversal and became agitated or unresponsive. In one patient, the procedure was completed with antegrade flow without protection; in the second patient, the situation was solved with intermittent rather than continuous carotid flow reversal, allowing antegrade cerebral flow in between carotid instrumentation maneuvers; and the third patient was converted to endarterectomy.

Bradycardia or hypotension, or both, requiring pharmacologic intervention occurred in response to carotid balloon dilatation in 24 cases (23%). Two patients became asystolic during balloon dilatation, one despite premedication with atropine, but both responded immediately to a precordial thump. Their troponin levels were not elevated. One of these patients required elective implantation of a pacemaker several days after CAS secondary to an underlying conduction block. Two patients became severely hypertensive in response to carotid balloon inflation. In addition, three patients became transiently unresponsive upon balloon inflation during stent postdilatation.

During follow-up at 3 to 40 months, all patients remained neurologically unchanged, and carotid ultrasound revealed that all but one carotid stent remained patent without recurrent stenosis. No cranial nerve injuries were identified. No stenoses or pseudoaneurysms were found in the common carotid puncture site. Five patients died during the follow-up period; three of causes not related to stroke, and the cause of death in two patients could not be determined. Seven patients were lost to follow-up. No patients progressed to renal failure during follow-up. One carotid stent was occluded at 1 month after intervention; no unusual circumstances occurred during the procedure. Life-table analysis of primary stent patency, on an intention-to-treat basis, was 95% at 40 months (Table II), and stroke-free patient survival was 91% (Fig 2).

Table II. Primary patency of 103 carotid stents based on intention to treat⁎

	Interval (mon)	Entering this interval	Exposed to risk	Lost to follow-up	Death	Failed stents, No.	Interval patency rate	Cumulative patency
	0-1	103	103	0	0	3	0.97	0.97
	1-4	100	99.5	1	0	0	1	0.97
	4-7	97	97	0	0	1	0.98	0.95
	7-10	94	94	0	1	0	1	0.95
	10-13	90	89.5	1	0	0	1	0.95
	13-16	86	85.5	1	0	0	1	0.95
	16-19	81	81	0	0	0	1	0.95
	19-22	77	77	0	2	0	1	0.95
	22-25	72	72	0	0	0	1	0.95
	25-28	67	67	0	0	0	1	0.95
	28-31	66	65	2	0	0	1	0.95
	31-34	62	61.5	1	1	0	1	0.95
	34-37	58	58	0	1	0	1	0.95
	37-40	53	52.5	1	0	0	1	0.95

⁎ Life-table analysis, on an intention-to-treat basis, of stent primary patency after transcervical stenting. The three failed stents from the first interval correspond to conversions to endarterectomy and not to stent occlusions.

View Large Image Download to PowerPoint

Fig 2. Life-table analysis of stroke-free patient survival after transcervical carotid stenting.

Discussion 

Cerebral embolization is a common event during carotid instrumentation maneuvers. This is demonstrated by the abundant embolic signals detected by transcranial Doppler (TCD) imaging in the middle cerebral artery during CAS without protection.15, 16 Not surprisingly, CAS without protection is associated with a 15% to 57% incidence of new, mostly asymptomatic, ipsilateral brain infarcts detected by postoperative magnetic resonance imaging (MRI).17, 18, 19 The use of distal carotid filters for cerebral protection during CAS decreases the incidence of cerebral embolization.20, 21, 22, 23 However, distal filters require crossing the carotid lesion before protection is in place, probably the most emboligenic maneuver necessary for the procedure.15, 16 This is perhaps the reason why the incidence of new MRI-detected brain infarction after CAS with distal filter protection is 26% to 41%, a similar rate to that occurring during CAS without cerebral protection.6, 7, 8

The clinical significance of the embolic phenomena during CAS is not well known, but a clear association exists between intraprocedural embolization and postprocedural brain infarction. With this in mind, it would be reasonable to expect that the use of a cerebral protection method that would eliminate intraprocedural embolization would likely produce a decrease in embolic-related neurologic complications after CAS.

Carotid flow reversal has shown to suppress cerebral embolic signals during CAS in vitro and in patients and could potentially eliminate embolic complications during CAS.9, 10, 11 The additional unique advantages of transcervical CAS with carotid flow reversal were extensively discussed in our previous report.12 The technical advantages of the procedure are that flow reversal is in place before the carotid lesion is crossed, it avoids the femoral access, and avoids arch and supra-aortic trunk instrumentation and any other unfavorable anatomy between the femoral arteries and the common carotids. In addition, it enhances carotid flow reversal throughout the procedure by using a larger caliber fistula and a shorter, lower resistance arteriovenous communication than that provided by a transfemoral flow reversal catheter. It is therefore plausible that the transcervical approach may produce more effective embolic suppression. The transcervical route also eliminates the embolic complications related to the performance of angiographic catheter manipulations necessary during transfemoral procedures, which may be associated with a significant incidence of new brain infarcts detected by MRI.24

The unique features of the transcervical approach for CAS perhaps explain the absence of major neurologic adverse events in our patients, despite being an initial experience with a newly devised technique. Our expanded experience with absence of major adverse events compares favorably with the combined death/major stroke rates for the recently published Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE),3 ACCULINK for Revascularization of Carotids in High-Risk Patients (ARCHer),4 and Endarterectomy Versus Angioplasty in Patients With Symptomatic Severe Carotid Stenosis (EVA-3S)5 carotid stenting trials, which were 1.8%, 3.6%, and 3.5%, respectively (Table III), a death/stroke rate statistically not different from that of their respective carotid endarterectomy cohorts. It is disappointing that these large, complex, and costly trials have eluded the analysis of cerebral embolization rates and of the incidence of new MRI-detected infarction of the brain. These unexplored study end points, directly related to the target organ of carotid intervention, would have been helpful in assessing the efficacy of distal filter protection in preventing brain infarction. Unfortunately, our study also lacks postoperative MRI evaluation.

Table III. Comparison table of results from this study with other studies

	Study	CAS cases, No.	Results at 30 days, %				3-year patency, %
			Death	Major stroke	Minor stroke	Death/any stroke
	SAPPHIRE	167	1.2	0.6	3.0	4.8	NR
	EVA-3S	265	0.8	2.7	6.1	9.6	NR
	ARCHer	581	2.1	1.5	4	6.9	97
	Present	103	0	0	2.0	2.0	97

CAS, carotid artery stenting; SAPPHIRE, Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy; EVA-3S, Endarterectomy Versus Angioplasty in Patients With Symptomatic Severe Carotid Stenosis; ARCHer, ACCULINK for Revascularization of Carotids in High-Risk Patients.

If we assume that most major strokes during CAS are the result of periprocedural embolization, we could speculate that the elimination of emboli during the procedure could possibly reduce the major stroke rate to levels significantly lower than those achieved with carotid endarterectomy, a currently elusive goal. Because of the already low stroke rates achieved by both procedures, a confirmation of this hypothesis would, unfortunately, require a very large comparison trial.

The influence of the type of stent used on the results of CAS has not been well investigated. The use of closed-cell stents has been suggested to be associated with a lower neurologic complication rate, a factor that could have potentially influenced our results in a positive manner.25

Our technical success rate was 97%, comparable with those reported in the above-mentioned major CAS trials with much larger experiences. We found no anatomic contraindications for the procedure in this series. However, we preoperatively evaluate the common carotid artery with duplex scanning, and we do not recommend the technique in patients with significant stenosis, calcification, or aneurysmal dilatation of the proximal common carotid artery.

The neurologic tolerance to carotid flow reversal was excellent in our experience. Only three of 72 patients (4%) who underwent the procedure under local anesthesia did not tolerate carotid flow reversal. All three patients had recently symptomatic (<2 months) carotid lesions, but only two of these three patients had contralateral ICA occlusion, accounting for a 20% intolerance to carotid flow reversal among 10 patients with contralateral ICA occlusion. This was an almost identical intolerance rate to that reported with carotid clamping during endarterectomy under local anesthesia in patients with contralateral ICA occlusion.26 Contralateral ICA occlusion with recent ipsilateral hemispheric symptoms, therefore, may be a relative contraindication to carotid flow reversal under local anesthesia. We have learned, however, that tolerance to flow reversal may be improved by reversing flow intermittently during short periods of time while carotid instrumentation is done and allowing antegrade flow in between maneuvers.

Our expanded experience has shown a decrease in the rate of hypotensive response to carotid balloon dilatation. This is likely secondary to the more liberal use of premedication with atropine in the later part of our experience. It is now our practice to premedicate all patients with atropine shortly before balloon dilatation.

The need for a surgical arterial cutdown could be considered a drawback of our technique; however, the wound complication rate was only 2%, comparable with the 2.6% access complication rate documented in the ARCHer trial.4 In the ARCHer trial, however, most femoral access site complications required blood transfusion, which was never the case with our technique.

In our expanded experience, the complication rate has remained low. Complications related to direct arterial puncture for carotid access decreased significantly with experience. During our first 12 interventions, direct carotid puncture caused four common carotid dissections: three were minor and corrected with the stent placement intended for CAS, and one required surgical correction. Only one dissection occurred during the last 91 cases, which was converted immediately to an uneventful carotid endarterectomy at the surgeon’s discretion. It appears, therefore, that the learning curve may be associated with technical complications mainly related to direct arterial puncture leading to common carotid dissection, without being associated with neurologic complications.

During the 3-year follow-up, our stent patency rate has been remarkably high, and no carotid reinterventions were needed. In a similar fashion, the stroke-free survival of our patients was comparable, if not superior, to that reported in the ARCHer trial.4 Our study is, admittedly, limited by the sample size and the lack of prospective exhaustive controls that are typical in large trials. Our results, however, suggest that carotid flow reversal can yield very low neurologic complication rates during CAS, potentially lower than those achieved with distal filter protection. If carotid flow reversal by the transcervical or any other route could eliminate the risk of cerebral embolization during CAS, it is plausible that it could achieve lower neurologic complication rates than those of CAS with distal filter protection. Carotid stenting could therefore become a safer procedure and possibly could achieve lower neurologic complication rates than carotid endarterectomy. These speculative hypotheses warrant a prospective controlled investigation.

Conclusion 

Carotid artery stenting is an evolving technique. We are convinced that with improved protection methods, optimization of stents and delivery systems, and refined technique, CAS will achieve better results than those currently available. Distal filters offer limited protection, and are far from the ideal method for cerebral protection during CAS.6, 7, 8 We need to continue to investigate alternative techniques and technologies to improve on the results of CAS.

Author contributions