SPACE: Carotid stenting falls short of noninferiority vs endarterectomy

Susan Jeffrey

October 06, 2006

Heidelberg, Germany - Results of a randomized trial of carotid stenting vs endarterectomy in symptomatic patients has failed to demonstrate noninferiority of the stenting procedure[1]. The trial, called the Stent-Supported Percutaneous Angioplasty of the Carotid Artery versus Endarterectomy (SPACE) study, actually showed a slightly higher rate of ipsilateral ischemic stroke and death at 30 days in the patients undergoing carotid stenting.

"The results of this trial do not justify the widespread use in the short term of carotid-artery stenting for treatment of carotid-artery stenoses," the SPACE investigators conclude in their report. "Patients should be informed that there are insufficient data available for long-term comparison of the two treatments."

The results were published in the October 7 issue of the Lancet.

An ongoing question

Carotid endarterectomy is still recommended by European and North American guidelines for the prevention of stroke or recurrent stroke in patients with symptomatic carotid stenosis, the authors, with corresponding author Dr Werner Hacke (University of Heidelberg, Germany), note. "However, with technological advancements in balloon-catheter and stent technology, carotid-artery stenting has become widely used, both in symptomatic and asymptomatic patients, despite an absence of evidence of a benefit of this treatment," they write.

In the SPACE trial, 1200 patients were randomly assigned within 180 days of a transient ischemic attack (TIA) or moderate ischemic stroke, defined as a modified Rankin score of <3, to undergo either carotid stenting (n=605) or carotid endarterectomy (n=595).

To be eligible, patients had to have carotid stenosis >70% on duplex ultrasound, to correspond to >50% of North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria, or >70% of the European Carotid Surgery Trial (ECST) criteria.

The trial was carried out in 35 centers in Germany, Austria, and Switzerland, with neurologists, vascular surgeons, and interventionalists participating from each center. A quality committee for each discipline enforced qualifications for each of these areas, and 10 centers were ultimately not allowed to participate because they failed to reach these standards. Use of cerebral protection devices was at the discretion of the participating center.

The primary end point for the trial was ipsilateral ischemic stroke or death at 30 days following the procedure; the trial was a noninferiority rather than a superiority trial, with a noninferiority margin defined as less than 2.5%, based on an event rate of 5%.

Final results showed that although the absolute difference in these events between the groups was 0.51%, the 90% confidence interval crossed zero, ranging from -1.89% to 2.91%.

"Even with the difference of 0.51%, the upper 90% confidence limited of this difference is 2.91%, which is greater than the predefined threshold of 2.5%," the authors write. "The p value of 0.09 for the calculated noninferiority test can be interpreted as there being a 9% chance of erroneously concluding that carotid-artery stenting is not inferior to carotid endarterectomy."

SPACE: Ipsilateral ischemic stroke and death with carotid stenting vs carotid endarterectomy
End point Carotid stenting, n (%) Carotid endarterectomy, n (%) Odds ratio (CAS vs CEA) (95% CI)
Primary end point 41 (6.84) 37 (6.34) 1.09 (0.69-1.72)
Ipsilateral ischemic stroke 39 (6.51) 30 (5.14) 1.26 (0.77-2.18)
Ipsilateral intracerebral bleeding 1 (0.17) 5 (0.86) 0.19 (0.004-1.74)
Death 4 (0.67) 5 (0.86) 0.78 (0.15-3.64)
CAS=carotid artery stenting; CEA=carotid endarterectomy

Per-protocol analysis showed no equivalent periprocedural complication rates, they note. "For all outcome analyses, there was a tendency toward better results in the carotid endarterectomy group, apart from death within 30 days and hemorrhagic stroke," consistency that supports the validity of their findings, they write.

The SPACE investigators plan to report long-term follow-up over time from the trial, they write. "These long-term results and the combined analysis with other large randomized controlled trials will enable us to draw more definite conclusions about the risks of the two treatments."

No systematic evidence

In an editorial, Dr A Ross Naylor (Leicester Royal Infirmary, Leicester, UK) discusses some of the important "take-home" messages of these results[2]. Among these: first, outcomes in this study were "considerably in excess of what is quoted in contemporary carotid endarterectomy or carotid angioplasty and stenting publications and registries," Dr. Naylor writes.

Rejection of a large number of potential collaborators because of their track record will raise questions about the generalizability of the results, he asserts. The nonuniform use of cerebral protection devices will lead some to say that this is why the trial did not show noninferiority of carotid stenting. "In fact, SPACE contradicts this viewpoint and has corroborated the interim data from the much-maligned EVA-3S study, which showed that cerebral protection devices were no panacea against procedural complications."

In SPACE, the 30-day rate of ipsilateral stroke and death with protection devices was 7.3%, vs 6.7% without, he notes.

"Accordingly, SPACE has provided surgeons and interventionalists with evidence to support their personal prejudices, and the next Cochrane review will almost certainly repeat its conclusion," Dr. Naylor writes. "Sadly, the main loser is the patient."

Despite SPACE being the largest trial to date, "one is left with the unavoidable conclusion that it was stopped prematurely," he writes; the power calculation was based on 1900 patients. Five of 8 randomized trials comparing these modalities have been stopped because of recruitment problems, Dr. Naylor notes. Two trials continue to recruit, CREST and ICSS.

"Contrary to what is generally assumed, no systematic evidence exists to support the preferential use of carotid endarterectomy over carotid angioplasty and stenting or vice versa," he concludes. "Please keep randomizing patients in ICSS and CREST."

Funding for the trial was provided by Federal Ministry of Education and Research, German Research Foundation, German Society of Neurology, German Society of Neuroradiology, German Radiological Society, as well as Boston Scientific, Guidant and Sanofi-Aventis. Dr. Hacke reported that he has received fees from Sanofi-Aventis; other coauthor disclosures are listed in the paper.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.