Appropriateness of learning curve for carotid artery stenting: an analysis of periprocedural complications

Objectives. Cerebral embolism is the first cause of neurologic complications of carotid artery stenting (CAS). A large debate has been raised to identify the caseload necessary for an appropriate learning curve before systematic use of CAS. This study examined (1) the timing of periprocedural complications during CAS and how these complications vary over time to identify factors that contribute to neurologic morbidity and (2) a sufficient number of procedures for adequate training. Methods: During 2001 to 2006, 627 CAS procedures with cerebral protection devices (CPD) were performed in a single vascular surgery center by a team including a vascular surgeon and an interventional radiologist. These represented 38% of a total of 1598 carotid revascularizations performed in the same interval. CAS procedures were divided into two groups according to time interval: the first period, 2001 to 2003, included 195 CAS procedures, and the second period, 2004 to 2006, included 432 CAS procedures. During each CAS procedure, five major steps were considered: phase 1, or the catheterization phase, included the passage of the aortic arch, catheterization of the target vessel, and introduction of a guiding catheter or sheath. Phase 2, or the crossing stenosis phase, included the placement of a CPD. Phase 3, or the stent ballooning phase, included predilation (when indicated), stent implantation, postdilation, and recovery of the protection system. Phase 4, or the early postinterventional phase, included the first 24 hours after leaving the catheterization table. Phase 5, or the late postinterventional phase, included the interval from the first postoperative day to 30 days. Results. At 30 days, 10 major strokes (2 of which were fatal) and 1 cardiac death occurred, for an overall major stroke/death rate of 1.75%. Furthermore, 18 minor strokes (2.9%) were recorded. By analyzing the occurrence of major strokes according to the three intraprocedural phases, four occurred in phase 1 and six in phase 3. All strokes but one were ischemic; six were ipsilateral, three were contralateral, and one was posterior. Minor strokes occurred prevalently after the procedure (11 in phase 4, 5 in phase 5, and 1 for phases 1 and 3). Comparing the first with the second interval of the study period, the 30-day major stroke and death rate decreased from 3.1% to 0.9% (P =.047), and the 30-day any stroke and death rate decreased from 8.2% to 2.7% (P = .005). According to multivariate analysis, study interval (hazard ratio, 3.68; 95% confidence interval, 1.49-9.01; P = .005) and age (hazard ratio, 1.06; 95% confidence interval, 1.00-1.12; P= .05) were significant predictors of stroke. Conclusions. A large proportion of major strokes (4/10) from CAS cannot be prevented by using CPD, because these strokes occur during catheterization (phase 1). This finding, together with the significant decrease in the overall stroke/death rate between the first and the last interval of the study period, enhances the importance of an appropriate learning curve that involves a caseload larger than that generally accepted for credentialing. The noticeable number of postprocedural cerebral embolizations leading to minor strokes and occurring in the early and late postinterventional phases (16/18) is likely due to factors less strictly related to the learning-curve effect, such as stent design and medical therapy. Moreover, expertise in selecting material and design of the stents according to different vessel morphology, in association with correct medical treatment, may be useful in reducing the number of minor strokes that occur in the later postinterventional phases of CAS.