Experts Continue to Grapple With Stenting or Endarterectomy for Carotid Stenosis


After more than a decade of clinical trials specifically comparing the two approaches, the choice of carotid artery stenting (CAS) or endarterectomy (CEA) for patients with asymptomatic or symptomatic carotid stenosis continues to be debated. And as several presentations at the VEITH symposium in New York City on November 17 made clear, that debate shows no signs of wrapping up in the near future.

Some speakers viewed long-term results from trials such as ACT I and ICASS, along with recent observational data, as evidence that stenting and surgery provide essentially equivalent outcomes, while others interpreted the findings as showing that CAS is inferior. Perhaps unsurprisingly, the audience—which consisted nearly entirely of vascular surgeons—favored the latter viewpoint in a show of hands.

Outcomes Comparable

Summing up key data in a formal debate, William Gray, MD, of Columbia University Medical Center (New York, NY), argued that prospective randomized and single-armed data support reasonable equivalence between CAS and CEA in many patients with standard or high surgical risk.

In CREST, which included asymptomatic patients, for example, the stenting and surgery groups had similar rates of any stroke, MI, or death within 30 days plus subsequent ipsilateral stroke (primary endpoint) and periprocedural complications through up to 4 years of follow-up.

Furthermore, he said, findings from prospective postmarketing studies of CAS have shown that 30-day rates of death or stroke meet or fall below the 3% and 6% thresholds established by guidelines for asymptomatic and symptomatic high-risk patients, respectively. Additionally, he reported, outcomes have improved from preapproval to postmarketing studies.

Gray acknowledged that multiple retrospective comparisons of CAS and CEA have shown better 30-day outcomes with surgery, but pointed out that all such assessments are “fatally flawed” due to selection and ascertainment bias. Compared with patients undergoing surgery, those treated with CAS generally have a higher level of risk and undergo more routine neurologic assessments that uncover more strokes, he said.

He concluded that CAS and CEA provide comparable outcomes in standard-risk patients, and that CAS is superior in patients with high surgical risk. But those conclusions are not static, he said, noting that proximal protection and next-generation devices have been shown to improve stenting outcomes in preliminary studies.

Long-term Outcomes Support Equivalence

The relative equivalence of CAS and CEA received support from 2 other presentations of long-term trial outcomes.

In the first, Jon Matsumura, MD, of the University of Wisconsin-Madison, reported 5-year results from the ACT I trial, which compared CAS with CEA in asymptomatic patients younger than 80 years with standard surgical risk.

The main results showed that the rate of any stroke, MI, or death within 30 days and ipsilateral stroke between 31 and 365 days (primary composite endpoint) was 3.8% with stenting and 3.4% with surgery, which met criteria for noninferiority. Other outcomes were similar, as well.

For the long-term outcomes, ipsilateral stroke from 30 days to 5 years, freedom from any stroke, and survival were all comparable between groups, whereas clinically driven reintervention was less frequent after stenting (1.6% vs 3.3%; P = .05).

In the second talk, Jonathan Beard, ChM, MEd, of Sheffield Teaching Hospitals NHS Foundation Trust (Sheffield, England), presented the previously reported long-term results from the ICSS trial, which including symptomatic patients.

Through a median follow-up of 4.2 years, the rate of stroke or death was higher in the CAS group (11.8% vs 7.2%), although the difference emerged in the first 120 days. After that point, rates were similar. Rates of ipsilateral stroke more than 30 days after treatment, fatal or disabling stroke, and restenosis did not differ by treatment type over the long term.

Beard said that the higher 30-day risk of stroke following CAS was equal to 1 extra stroke for every 156 patient-years of follow-up. “But you have to question whether it’s clinically that relevant, especially as carotid endarterectomy has a higher risk of cranial nerve injury, which can result in a severe disability, and bleeding.”

He added that there were no differences between the CAS and CEA arms in long-term functional outcome, quality of life, or cost-effectiveness.

“Patients choosing CAS can be reassured about the long-term benefit, but there are differences in procedural outcomes and those need to be explained clearly and honestly to the patients,” he said.

Surgery Tops Stenting

Other speakers were less convinced of the similarity in outcomes with CAS and CEA.

In the debate with Gray, A. Ross Naylor, MD, of Leicester Royal Infirmary (Leicester, England), started his talk by highlighting the discrepancy in outcomes between trials of CEA and real-world data, with much higher rates of death or stroke outside of the trial setting. Real-world analyses have been similarly unkind to CAS, he continued.

A recent study by his group in the European Journal of Vascular and Endovascular Surgery examined rates of death or stroke following CAS or CEA in 21 registries spanning 2007 to 2014; it included more than 1.5 million procedures.

In average-risk asymptomatic patients, risks of death or stroke were similar after CAS and CEA in 24% of the registries but higher after CAS in the rest (either significantly or with no statistical comparison). Risk exceeded the 3% threshold in only 1 registry for CEA but in 9 for CAS.

Moving to symptomatic patients with average risk, risks of death or stroke were similar after CAS and CEA in 11% of registries, with higher risks after CAS in the rest. Risk was above the 6% threshold in only 1 of 18 registries for CEA but in 13 for CAS.

In both asymptomatic and symptomatic patients, death or stroke rates did not decline over time for either CAS or CEA.

“In the real world, stenting is being performed with significantly higher procedural risks, and most importantly, with procedural risks well in excess of the American Heart Association guidelines, especially in symptomatic patients,” Naylor said.

Likewise, in a rebuttal to Beard’s talk, Anne Abbott, MBBS, PhD, of Monash University (Melbourne, Australia), argued that the late ICSS results do not, in fact, show the equivalence of CAS and CEA for symptomatic carotid stenosis.

She said that the focus on similar rates of fatal or disabling strokes incorrectly discounts the fact that the rate of any stroke was 1.7 times higher with CAS. All strokes are disabling to some degree and matter, she said.

The trial also showed that rates of any disability were similar in the CAS and CEA arms of the trial up to 5 years. However, Abbott said, looking at any disability masks the effect of stroke-related disability caused by stenting.

And finally, she disputed the notion that CAS-caused stroke is offset by CEA-caused MI. In an analysis of ICSS and 6 other trials, she noted, stroke was 4.5 times more likely than MI at 30 days.

She concluded that CEA is superior to CAS for symptomatic carotid stenosis, while pointing out that the evidence of a benefit over medical treatment alone is outdated and limited to small subgroups.

Routine CEA or CAS Not Needed for Asymptomatic Patients

In another talk, J. David Spence, MD, of Robarts Research Institute (London, Canada), highlighted data showing that results of both CAS and CEA are worse in the real world than in clinical trials. As such, it’s important to remember the conditions that need to be met before either approach is considered.

CEA is indicated for symptomatic severe carotid stenosis, whereas CAS is indicated for symptomatic severe stenosis with special features, including high bifurcation, previous radiation, repeat surgery, high medical risk, and younger patient age, he said.

In asymptomatic patients, he said, CEA or CAS should be used only for patients with high-risk plaques, indicated by microemboli on transcranial Doppler, intraplaque hemorrhage on MRI, hot plaques on PET/CT, or reduced cerebrovascular reserve.

Intensive medical therapy is needed for all patients with carotid stenosis and is the best treatment for 90% of patients with asymptomatic disease, Spence concluded.  “Routine stenting or endarterectomy is not warranted for asymptomatic stenosis.”
 

Sources
  • Matsumura JS. Update on the ACT I trial comparing CAS and CEA in patients with high-grade asymptomatic carotid stenosis. Presented at: VEITHsymposium; November 17, 2015; New York, NY.

  • Gray WA. DEBATE: CAS is equivalent to CEA in many circumstances. Presented at: VEITHsymposium; November 17, 2015; New York, NY.

  • Naylor R. DEBATE: Not so: lessons learned from US datasets on outcomes after CEA and CAS show CAS results are substantially worse than in CREST: the brighter the light the darker the shadow. Presented at: VEITHsymposium; November 17, 2015; New York, NY.

  • Beard JD. DEBATE: Late results of ICSS trial show CAS and CEA to be equivalent in patients with symptomatic carotid stenosis. Presented at: VEITHsymposium; November 17, 2015; New York, NY.

  • Abbott AL. DEBATE: Late results of ICSS do not show equivalence of CAS and CEA: CEA still wins for symptomatic carotid stenosis. Presented at: VEITHsymposium; November 17, 2015; New York, NY.

  • Spence JD. Stroke and death rates (early and late) after CAS are higher than the randomized controlled trials (RCTs) would indicate: therefore, we should be less aggressive with our use of CAS, especially in the elderly and asymptomatic patients. Presented at: VEITHsymposium; November 17, 2015; New York, NY.

Disclosures
  • ACT I was sponsored by Abbott Vascular.
  • Matsumura reports receiving research grant support in the past 3 years from Abbott, Cook, Covidien, Endologix, and Gore.
  • Abbott reports that her presentation was supported by the Bupa Health Foundation.
  • Gray and Spence did not make statements regarding relevant conflicts of interest.
  • Naylor reports no relevant conflicts of interest.

We Recommend

Comments