Assessing the impact of distal protection filter design characteristics on 30-day outcomes of carotid artery stenting procedures.

ABSTRACT

OBJECTIVE: This study aims to review retrospectively the records of patients treated with carotid artery stenting (CAS) to investigate the potential correlations between clinical variables, distal protection filter (DPF) type and characteristics, and 30-day peri-/postprocedural outcomes. METHODS: This is a multicenter, single-arm, nonrandomized retrospective study of patients who underwent filter-protected CAS in the Pittsburgh, Pennsylvania, region between July 2000 and May 2011. Analysis of peri-/postprocedural complications included myocardial infarction, transient ischemic attacks (TIA), stroke, death, and a composition of all adverse events (AEs). Filter characteristics for Accunet (Abbott Vascular, Santa Clara, Calif; n = 429 [58.8%]), Angioguard (Cordis Endovascular, Miami Lakes, Fla; n = 114 [15.6%]), FilterWire (Boston Scientific, Natick, Mass; n = 113 [15.5%]), Spider (ev3 Endovascular, Plymouth, Minn; n = 45 [6.2%]), and Emboshield (Abbott Vascular; n = 24 [3.3%]) were previously determined in vitro and were used to find correlations with CAS procedural outcomes. Both univariate and multivariate analyses were performed, as well as goodness-of-fit tests to find multivariate correlations with procedural outcomes. RESULTS: In total, 731 CAS procedures using six different DPFs were analyzed. Peri-/postprocedural AEs included 19 TIAs (2.6%), 38 strokes (5.2%), one myocardial infarction (0.1%), 19 deaths (3.6%), and a total of 61 patients with complications (8.3%). Univariate analysis for filter design characteristics showed that the composite of AE was negatively associated with both vascular resistance (P = .01) and eccentricity (P = .02) and was positively associated with porosity (P = .0007), number of pores (P = .005), and pore density (P = .001). Multivariate analysis and the goodness-of-fit test revealed that patients with a history of congestive heart failure, stroke, and TIA (each with odds ratio >1) led to a good-fit model P value of .72 for peri-/postprocedural AEs. Multivariate analysis was inconclusive for all filter design characteristics. CONCLUSIONS: The following filter design characteristics are independently significant for minimizing peri-/postprocedural AEs higher vascular resistance, concentric in shape, greater capture efficiency, lower porosity, lower number of pores, and lower pore density. Lower porosity and smaller wall apposition were also found to be independently significant for minimization of peri-/postprocedural TIAs. This information can be used when considering the desirable design characteristics of future DPFs.).