Carotid lesion characteristics are major factors affecting stroke and death in CREST


CREST (Carotid Revascularization Endarterectomy vs Stent Trial) has demonstrated a higher periprocedural stroke and death rate among patients randomised to carotid artery stenting (4.4%) than carotid endarterectomy (2.3%, p=0.005). In previous publications, in an attempt to identify high-risk groups, it has been noted that patients who were older than 70 years did better with carotid endarterectomy than with angioplasty and stenting (p=0.02). In another publication it was noted that women undergoing carotid stenting had a higher event rate than women undergoing carotid endarterectomy.

The results of the new analysis were presented by Wesley S Moore, professor and chief, Emeritus Division of Vascular Surgery UCLA Medical Center, Los Angeles, USA, on behalf of the CREST investigators at the Vascular Annual Meeting (17–20 June, Chicago, USA).

Moore commented: “The question that has been raised is whether age and gender are surrogates for other factors that we might be able to identify. In this analysis, we sought factors (patient and/or arterial characteristics) that affect the treatment differences between stenting and endarterectomy. We also aimed to identify a subgroup of patients where stenting and endarterectomy have equivalent periprocedural stroke and death risk.”

Patient and arterial plaque characteristics were assessed as possible effect carotid artery stenting and endarterectomy result modifiers using a logistic regression model.

The CREST trial enrolled 2,502 patients. Preprocedure angiography was available in 438 (35%) out of 1,240 endarterectomy patients and in all stenting patients (1,262). Endarterectomy patients with available angiography did not differ from stenting patients with respect to age, gender, symptomatic status, smoking history, arrhythmia or LVH.

The results, Moore said, showed that there was increased risk of stroke and death with stenting compared to endarterectomy in the following three areas: long lesions (median >12.85mm), dyssynchronous or sequential lesions, and lesions distal to the carotid bulb. In patients with long lesions, the stroke and death rate with endarterectomy was 1.9% and with stenting 6.1% (odds ratio [OR] 3.45). In patients with sequential lesions it was 0.7% with endarterectomy and 5.8% with stenting (OR 9.21). Among patients with short sequential or remote lesions, the stroke and death rate was 1.6% with endarterectomy and 5.2% with stenting (OR 3.55). And finally, for patients with long lesions that were both remote and sequential, the stroke and death rate was 0% with endarterectomy and 6.3% with stenting (OR infinite in favour of endarterectomy).

Moore noted that with this information it was possible to populate a list of conditions that represent high risk for carotid artery stenting. He added, “From CREST we have learnt that long lesions, sequential lesions (perhaps those at the bulb and more distally) and distal lesions constitute high risk for stenting. From other trials we can also add to this list type 3 aortic arch, atherosclerotic aortic arch, tortuosity of the internal carotid artery, circumferential calcification and ulcerative lesions.”

Together, Moore said, these lesion characteristics represent 67% of the CREST population.


In conclusion, he stated, “For patients with these factors we strongly recommend that carotid endarterectomy be employed rather than stenting. However, in the absence of these high-risk characteristics, stenting appears to be as safe as endarterectomy with regard to periprocedural risk of stroke and death.”