Duplex scan surveillance after carotid angioplasty and stenting: A rational definition of stent stenosis

Methods 

Patients 

A total of 111 patients (93 men, 18 women) with a mean age of 64 years (range, 54 to 83 years) underwent 114 CAS procedures for symptomatic (n = 62) or asymptomatic (n = 52) carotid occlusive disease. Of these, 86 patients (75%) had treatment of severe ICA atherosclerotic stenosis, 28 were treated for >75% DR recurrent stenosis after CEA, and three had staged CAS procedures for bilateral >75% DR CEA-site stenosis:

Eligibility for CAS was determined by Centers for Medicare & Medicaid Services (CMS) guidelines for CAS coverage, including voluntary randomization in on-going multicenter clinical CAS trials and high-risk operative patients.10 Eighteen patients were enrolled in the Carotid Revascularization Endarterectomy versus Stent Trial (CREST), one patient in the Acculink for Revascularization of Carotids in High-Risk Patients Trial (ARCHeR), and the remaining 92 patients were judged to be high-risk for CEA (neck irradiation, prior CEA, severe cardiac/pulmonary disease) and had aortic arch and ICA anatomy suitable for CAS (Table I).

Table I. Medical and anatomic conditions in 92 patents judged to be high-risk for surgical carotid endarterectomy

	Variables	Patients (n)
	Medical comorbidities
	Cardiac disease	31
	Ischemic cardiomyopathy⁎	12
	Angina†	12
	Cardiac surgery‡	4
	Recent MI (<30 days)	3
	COPD	10
	Oxygen-dependent	6
	Steroid-dependent	2
	FEV1 <30%	2
	Anatomic conditions	57
	Prior ipsilateral CEA	28
	Prior neck irradiation	11
	High cervical ICA lesion	10
	Contralateral ICA occlusion	8
	Prior radical neck dissection	1
	Tracheostomy	1

COPD, Chronic obstructive pulmonary disease; FEV1, Forced expiratory volume in 1 second; CEA, carotid endarterectomy; ICA, internal carotid artery.

⁎ American Heart Association stage III or IV; ejection fraction <30%. † American Heart Association class III or IV. ‡ Anticipated or performed <30 days.

Patients who underwent CAS but did not complete the minimum 12-month protocol of DUS surveillance were excluded from review.

Surveillance after carotid angioplasty and stenting 

All patients undergoing CAS procedures were evaluated ≤1 month of the procedure, with most having a carotid DUS scan both before discharge and at 1 month after CAS. Carotid testing was performed in an accredited (Intersocietal Commission on Accreditation of Vascular Laboratories) testing facility. A 60° Doppler angle of insonation was used when possible to record midstream velocity spectra (PSV, end-diastolic velocity [EDV]) from the common carotid artery (CCA), along the stent length, and in the ICA distal to the stent. Power Doppler imaging of the stent and adjacent artery segments was performed to assess caliber and sites of maximum stenosis. The highest PSV value recorded from the stent was used with the proximal CCA value to calculate the ICAstent/CCA PSV ratio, or proximal stent PSV to calculate the PSVStent ratio with a value >2 indicating stenosis. B-mode imaging was also used to record transverse and anteroposterior stent diameters in the proximal, middle, and distal stent regions. Interpretation criteria used to estimate stenosis severity after CAS classified stenosis into four categories: <50% DR, 50% to 75% DR, >75% DR, and occlusion (Table II).

Table II. University of South Florida duplex ultrasound scan criteria for grading carotid stent stenosis

	Stenosis category (DR)	PSV (cm/s)	PSV ratio	EDV (cm/s)	Color/power Doppler scan imaging results
	<50% (none)	<150	<2	NA	No or minimal stent lumen reduction
	50%-75% (moderate)	>150	>2	<125	Turbulent flow, stent lumen reduction present
	>75% (severe)	>300	>4	>125	High-grade stent stenosis, damping of distal ICA spectral waveform
	Occlusion	NA	NA	NA	No stent flow visualized

DR, Diameter reduction; PSV, peak systolic velocity; EDV, end diastolic velocity; NA, not applicable.

DUS surveillance was performed at 6-month intervals after the initial 1-month evaluation. A shorter 3-month interval between scans was performed in patients with >50% DR residual stenosis and when stent stenosis progression from <50% DR to >50% DR was detected, and if no further progression occurred, the surveillance interval was increased to 6 months. When high-grade, >75% velocity spectra were identified, angiographic imaging was recommended with consideration for intervention if a high-grade stent stenosis was confirmed.

Results 

Periprocedural outcomes 

No procedural strokes, death, or stent occlusions occurred. One patient required intraprocedural thrombolysis for middle cerebral artery occlusion. Initial duplex testing identified normal (PSV <150 cm/s) velocity spectra at 90 CAS sites (79%) (PSV, 94 ± 24 cm/s; PSV ratio, 1.2 ± 0.3; Fig 1). Velocity spectra in the 50% to 75% DR category (PSV, 183 ± 34 cm/s; PSV ratio, 2.5 ± 0.5) were recorded from 23 CAS sites (20%), and one stent had a PSV of 355 cm/s and an EDV of 126 cm/s, indicating a high-grade residual stenosis. This patient had been treated for an asymptomatic 85% DR CEA-site stenosis, and balloon angioplasty was not performed after stent deployment owing to difficult aortic arch anatomy and loss of CCA catheter access. A completion angiogram demonstrated a 30% residual stent stenosis. A follow-up DUS scan at 4 months documented asymptomatic stenosis progression (PSV, 550 cm/s; EDV, 200 cm/s; PSV ratio, 11) and surgical repair with stent explant and vein patch angioplasty was performed 5 months after CAS.

View Large Image Download to PowerPoint

Fig 1. Classification by percentage of diameter reduction (DR) of carotid stent stenosis at 1 month, and 1 year, maximum diameter reduction during surveillance, and diameter reduction at the time of last follow-up, including intervention for >75% DR. CAS, carotid angioplasty and stenting.

Carotid stent surveillance 

During the mean 33-month follow-up period (range, 12 to 78 months), three patients experienced ipsilateral, nondisabling, reversible neurologic events at 30, 45, and 120 days after CAS; none was associated with severe stent stenosis (<50% DR in 2; 50% to 75% DR in 1) or cerebral infarction on computed tomography or magnetic resonance imaging.

Compared with the initial DUS testing, serial scans detected progressive CAS stenosis from <50% DR (PSV, 97 ± 25 cm/s) to 50% to 75% DR (PSV, 217 ± 38 cm/s) at 21 (23%) of 90 sites, including three sites (4%) that subsequently progressed to >75% DR in-stent stenosis (Table III). Velocity spectra indicating stenosis regression or progression was recorded from 23 CAS sites with 50% to 75% DR stenosis on initial scanning, including nine sites with stenosis regression from 50% to 75% DR (PSV, 195 ± 42 cm/s) to <50% DR (PSV, 115 ± 13 cm/s), and two sites with progression to >75% DR in-stent stenosis. The mean time of CAS site regression (n = 9) was 9 ± 6 months compared with 14 ± 10 months for stenosis progression (n = 23). Of note, stent stenosis that developed after initial normal (<50% DR category) DUS testing was not observed to regress on subsequent scans.

Table III. Summary of cohort carotid stent surveillance: initial and last recorded duplex study according to in-stent diameter reduction

	DR	Initial study, n (%)(n=114)	No change DR on last study, n (%)	Regression of CAS stenosis, n (%)	Progression to <50% DR, n (%)	Progression to high-grade stenosis, n (%)
	<50%	90(78.9)	69(60.5)	NA	18⁎(15.8)	3⁎(2.6)
	≥50%	23(20.2)	12(10.5)	9†(7.9)	NA	2⁎(1.8)
	>75%	1(0.9)	1(0.9)	0	NA	NA

DR, Diameter reduction; NA, not applicable. Reintervention: radiation atherosclerosis (n = 2), recurrent carotid endarterectomy stenosis (n = 1), primary atherosclerotic stenosis (n = 2).

⁎ Mean time to progression of carotid artery stent stenosis 14 ± 10 months. † Mean time to regression of carotid artery stent stenosis 9 ± 7 months.

By life-table analysis, freedom from DUS-detected >50% DR stenosis was 79% at 1 month, 78% at 6 months, 76% at 1 year, and 67% at 4 years, and freedom from >75% DR stenosis was 97% at 1 year and 95% at 4 years (Fig 2). No stent occlusion was identified. The risk for developing a persistent >50% DR stent stenosis was similar after treatment for CEA-site stenosis (9 [32%] of 28), atherosclerotic ICA stenosis in an irradiated neck (3 [33%] of 9), and ICA stenosis in a nonirradiated neck (23 [31%] of 75).

View Large Image Download to PowerPoint

Fig 2. Freedom from carotid stent stenosis of >50% (triangles) or >75% (squares) diameter reduction (DR) based on duplex ultrasound surveillance using life-table analysis.

The yield of DUS surveillance for detection of high-grade CAS stenosis (PSV, 437 ± 98 cm/s [range, 301 to 578 cm/s]; EDV, 167 ± 29 cm/s [range, 126 to 201 cm/s]) was 5% (6/114 CAS); all associated with asymptomatic stenosis progression. Diagnostic biplane angiography confirmed a ≥75% DR in-stent stenosis in all six patients with DUS-detected high-grade CAS stenosis. The angiographic findings resulted in reinterventions consisting of balloon angioplasty (n = 3), stent angioplasty (n = 2), or surgical repair (n = 1). DUS testing after reintervention confirmed <50% DR at five sites and 50% to 75% DR (PSV, 181 cm/s; EDV, 87 cm/s) after balloon angioplasty in one patient. No further stenosis progression has been observed to date in this group.

Beyond 30 days, no patient with >50% DR stent stenosis on initial testing or subsequent DUS surveillance developed ipsilateral stroke. Six patients died during follow-up of cardiovascular or pulmonary disease.

Discussion 

The efficacy of CAS for stroke prevention is currently under investigation in the National Institutes of Health sponsored, multicenter, randomized CREST clinical trial. The completed Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) clinical trial, which compared outcomes between CEA and CAS with embolic protection in high-risk patients, found equivalent stroke rates, but the incidence of reintervention was higher (P = .04) after CEA (4.3%) than after CAS (0.6%).11 For CAS to demonstrate equivalent stroke prevention compared with medical management or CEA, low procedural morbidity (stroke, death) coupled with durability and a low incidence of restenosis requiring reintervention will need to be achieved.

Our vascular group has recommended a policy of DUS surveillance after carotid repair and reoperation when high-grade stenosis is identified.9 Patient outcomes after CEA and now CAS interventions have shown the yield of DUS surveillance to be similar: 5.9% intervention rate after CEA primarily for treatment of contralateral ICA disease progression compared with 10% after CAS with an equivalent intervention rate for high-grade CAS stenosis (5%) or contralateral ICA stenosis (5%). After surgical or endovascular carotid intervention, stenosis progression typically occurred without the appearance of neurologic events. Intervention for asymptomatic high-grade stenosis (>75% to 80%) was accomplished successfully with minimal procedural morbidity and was associated with low long-term stroke rates of <1% per year. After CAS, progression of ICA stent stenosis was 2.5 times more common (24 sites vs 9 sites) than regression and typically occurred >6 months of the procedure (mean, 14 months).9 Carotid occlusion was not observed after CEA or CAS partly because intervention for high-grade stenosis was performed.

Criteria for the interpretation DUS-detected stenosis after CAS are in evolution, including defining the threshold for reintervention. Moderate stenosis (>50% DR) involving the CAS repaired was detected in approximately 20% patients on initial testing and during surveillance. This incidence of moderate stenosis is similar to other reports using similar velocity spectra criteria, including Zhou et al12 (16% with ICAstent/CCA >3.2) and Lal et al8 (20% with PSV >150 cm/s, ICAstent/CCA >2.2).

The development of high-grade CAS stenosis is of concern because further progression may lead to occlusion and stroke. To date, most reports on high-grade stenosis of ≥70% have found the patients to be asymptomatic, and the need for reangioplasty or a more complicated surgical bypass or repair is controversial.12 Our criteria for >75% in-stent stenosis includes the combined velocity spectra criteria of PSV >300, EDV >125, and ICAstent/CCA >4, but the clinical decision to proceed with reintervention should be based on angiographic verification of stenosis severity, anatomic features of the lesions, patient symptoms, patency status of the contralateral ICA, and the anatomy of the circle of Willis—all factors influencing the risk vs benefit for reintervention.

Because no patient in our series developed a symptomatic high-grade stenosis, the decision to proceed with an endovascular or surgical repair was made on duplex findings of progressive in-stent stenosis, its verification by arteriography, and a perceived risk that progression could result in stroke. Endovascular intervention for high-grade stent stenosis was successful in restoring functional patency and was associated with a low incidence of recurrence. Repeat intervention may be higher in patients initially treated for CEA-site stenosis, but this risk factor was not found in our patient series.12, 13 An EDV threshold velocity >125 to 140 cm/s together with color/power Doppler imaging criteria of severe <2 mm lumen reduction appear to accurately predict a >75% to 80% DR stent stenosis.

DUS scanning can readily image carotid bifurcation stents, and serial testing can identify stent-related abnormalities, including thrombosis, in-stent stenosis, stent deformity, lack of apposition to artery wall, and migration. Stent deployment alters wall compliance of the covered carotid artery segment, producing a stiffer conduit and, theoretically, an increase in PSV in the stent, but the PSV ratio along the stent length should be <2. Blood flow patterns within a nonstenotic stent are nondisturbed except at proximal and distal stent orifices where diameter/compliance mismatch is present. Stent structure should contact the artery wall and plaque, because the incidence of stent failure and migration is higher when ultrasound imaging demonstrates poor wall apposition.

Serial ultrasound imaging has shown that both positive (stent expansion) and negative (stent lumen reduction due to myointimal hyperplasia) remodeling of the treated stenotic ICA occurs.14 The self-expanding stent diameter increased for several months after deployment, most evident in the middle stent region, but was negated by the presence of calcified plaque. In-stent neointimal thickening is a common finding, and its thickness increases for up to 12 months and usually stabilizes thereafter.

Thus, serial DUS testing with velocity spectra recordings indicating stent stenosis regression or progression is not surprising. A carefully conducted study from the vascular group in Vienna, Austria found the arterial remodeling after CAS most commonly produced a PSV increase indicating a dominance of negative remodeling secondary to myointimal proliferation.14 Progression of in-stent stenosis requiring intervention was uncommon (3.4%) using DUS interpretation criteria similar to this study.

When color Doppler imaging of the CAS segment demonstrates no stent lumen reduction, a maximum PSV <150 cm/s, and PSVstent ratio <2, assignment to <50% DR disease category is appropriate. These DUS findings indicate a nonstenotic or minimally stenotic ICA segment and are associated with a low subsequent risk for occlusion or stroke. Data from the CREST core DUS reading center found that 85% of initial DUS studies had a PSV <125 cm/s and 93% had a PSV <150 cm/s. In our experience, approximately three quarters of CAS sites were in this category initially and throughout the 3-year surveillance period. For patent CAS sites with PSV >150 cm/s, PSVstent ratio >2, two stenosis severity categories of 50% to 75% DR and >75% DR are useful to track stenosis progression or regression, with the >75% DR category indicating high-grade stenosis and signaling the threshold for additional patient evaluation including possible intervention.

One limitation of this review was the absence of structured angiographic verification of DR and stent stenosis. Although omitted in this study, angiographic verifications have previously confirmed the utility of DUS for predicting recurrent stenosis and failing vascular interventions in a variety of vascular beds. Routine angiography in asymptomatic patients with minimal or moderate restenosis is difficult to justify in a review of this type, especially because only asymptomatic high-grade (>75%) carotid occlusive disease is thought to be clinically significant. Our experience and other published reports support these DUS interpretation criteria of moderate and high-grade stent stenosis. An in-stent stenosis with severe hemodynamic abnormalities (PSV >300, EDV 125 to 140 cm/s, and PSVstent ratio >4) is likely to have >75% DR lumen reduction on angiographic imaging.

Clearly, DUS scanning is an accurate modality for identifying stent stenosis but does the traditional label of DUS overestimation of stenosis apply to CAS surveillance? The diagnostic acumen for carotid DUS imaging is highest in confirming the minimal (<50%) diameter reduction, stent patency or stent occlusion; however, stents demonstrating findings of intimal hyperplasia on power Doppler imaging or 50% to 75% diameter reduction by velocity criteria may represent a group that may be prone to progression of in-stent stenosis. Therefore, grading >50% stent stenosis may be prone to overestimation error, and confirmatory imaging with angiography should be considered if carotid reintervention is deemed gainful for symptomatic or asymptomatic patients.

Overestimation of high-grade stent stenosis was not a common event in this series. Both asymptomatic patients with DUS studies demonstrating >75% stent stenosis and symptomatic patients after CAS had diagnostic imaging studies that verified DUS findings to be accurate in directing the need for carotid reintervention. As a result of this review, we rely on DUS surveillance to provide a safe and effective noninvasive diagnostic tool to predict high-grade restenosis after CAS, allowing us to limit unnecessary exposure to radiation and contrast mediums in patients who are not likely to benefit from carotid reintervention.

Testing intervals of 6 months are sufficient to detect CAS site stenosis and monitor 50% to 75% stenotic lesions for progression. However, surveillance is also important to detect contralateral ICA stenosis progression. Imaging the CAS site ≤1 month is useful to exclude residual stenosis and reconfirm the severity of contralateral disease. If DUS testing confirms <50% ICA stenosis bilateral beyond the first 18 months, an annual scan is adequate for disease monitoring. Surveillance every 6 months is recommended in patients with >50% DR ipsilateral or contralateral ICA stenosis. The development of hemispheric symptoms in the presence of >50% DR ICA or CAS stenosis, or asymptomatic disease progression to a high-grade stenosis (>75% to 80% DR, EDV >140 cm/s), should prompt a recommendation of surgical or endovascular (stent-assisted angioplasty) intervention in appropriate patients.

Conclusion 

DUS surveillance after CAS identified a 5% procedural failure rate due to the development of high-grade in-stent stenosis, a higher clinical yield than CEA surveillance. Both progression and regression of stent stenosis severity was observed on serial testing, but 70% of CAS sites demonstrated velocity spectra consistent with <50% DR. Contralateral disease progression remains a risk factor, with a 5% intervention rate after both CEA and CAS using similar velocity spectra criteria indicating >75% DR stenosis. Our policy of DUS surveillance and reintervention for high-grade stenosis was associated with sustained stent patency and infrequent neurologic events.

Author contributions