Influence of multiple stents on periprocedural stroke after carotid artery stenting in the Carotid Revascularization Endarterectomy versus Stent Trial (CREST)

Background: In the Carotid Revascularization Endarterectomy versus Stent Trial (CREST), carotid artery atherosclerotic lesion length and nature of the lesions were important factors that predicted the observed difference in stroke rates between carotid endarterectomy and carotid artery stenting (CAS). Additional patient-related factors influencing CAS outcomes in CREST included age and symptomatic status. The importance of the operator's proficiency and its influence on periprocedural complications have not been well defined. We evaluated data from CREST to determine the impact of use of multiple stents, which we speculate may be related to technical proficiency. Methods: CREST includes CAS performed for symptomatic ≥50% carotid stenosis and asymptomatic ≥70% stenosis. Both symptomatic and asymptomatic patients were enrolled in the trial and in the lead-in registry. Data from patients enrolled in the CREST registry and randomized trial from 2000 to 2008 were reviewed for patient- and lesion-related characteristics along with number of stents deployed. The occurrence of 30-day stroke and demographic and clinical features were recorded. Odds ratios for 30-day stroke associated with the use of multiple stents were calculated in univariate analysis and on multivariable analysis after adjustment for demographics (age, sex, symptomatic status), lesion characteristics (length, ulceration, eccentric, percentage stenosis), and risk factors (diabetes, hypertension, dyslipidemia, and smoking). Results: The registry (n = 1531) and trial (n = 1121) enrolled 2652 patients undergoing CAS. The mean age was 69 years; 36% were women, and 38% were symptomatic. The mean diameter stenosis was 78%, and the mean lesion length was 18 mm (±standard deviation, 8 mm). Risk factors included hypertension (85%), diabetes (32%), dyslipidemia (84%), and smoking (23%). All patients received Acculink stents (Abbott Vascular, Abbott Park, Ill) that were 20, 30, or 40 mm in length (straight or tapered) and Accunet (Abbot Vascular) embolic protection when possible. Most patients received one stent (n = 2545), whereas 98 patients received two stents and 9 patients received three stents (P <.001) to treat the lesion. Patients receiving more than one stent were older (P =.01) but did not differ in other demographic or risk factors. Strokes occurred in 118 (4.5%) of all CAS procedures, in 102 (4%) with the use of one stent, and in 16 (15%) with the use of two or three stents. After adjustment for demographics, lesion characteristics, and risk factors, the use of more than one stent resulted in 2.90 odds (95% confidence interval, 1.49-5.64) for a stroke. Conclusions: Although we know that lesion characteristics (length, ulceration) play an important role in CAS outcomes, in this early experience with carotid stenting, a significant and independent relationship existed between the number of stents used and procedural risk of CAS. We postulate that this was an indicator of the operator's inexperience with the procedure.