Predictive factors of recurrent endovascular intervention for cephalic arch stenosis after percutaneous transluminal angioplasty.

OBJECTIVE: Cephalic arch stenosis (CAS) is an important and recurring problem arising in hemodialysis patients because of the requirement for repeated interventions to maintain patency. The aim of this study was to determine predictive factors for recurrence of CAS after successful angioplasty.

METHODS: A retrospective, case-control study was conducted at two ambulatory vascular access (VA) centers. All patients with a dysfunctional VA referred for an angiographic procedure and with a documented CAS as evidenced during the endovascular intervention (EI) between January 1, 2013, and December 31, 2015, were enrolled; 15 patients in whom an efficacious intervention was not possible were excluded. The study thus concerned 375 EIs using percutaneous transluminal angioplasty without stent placement on 241 VAs for CAS (9% of all procedures performed) during a 3-year period. Patients were compared regarding the absence (group 1; n = 181) or presence (group 2; n = 60) of recurrent CAS. We defined recurrence as that which occurred within 180 days of the previous successful EI for CAS. Any CAS diagnosed and treated >180 days after a previous one was considered a novel CAS and not a recurrence. Multivariate analysis was performed to determine variables independently associated with recurrence of CAS. Kaplan-Meier analysis was performed for determination of primary and assisted primary patency in this population.

RESULTS: The recurrence rate of CAS was high (25%). Patients in both groups had similar demographic characteristics, time on hemodialysis, and mean dialysis dose and access flow rate at referral (P > .05). Multivariate analysis provided a significant discriminatory influence pertaining to diabetes (hazard ratio [HR], 2.054; 95% confidence interval [CI], 1.22-3.46; P = .007), residual stenosis even though it was <30% (HR, 1.86; 95% CI, 1.005-3.439; P = .048), and the finding of an isolated CAS lesion (HR, 0.445; 95% CI, 0.219-0.905; P = .025) in comparing group 1 and group 2. All other variables lost statistical significance on multivariate analysis. Primary patency at 6 months was 72%, increasing to an assisted primary patency of 89% at 6 months. The median durations of primary patency and assisted primary patency were 9.5 months and 15.6 months, respectively.

CONCLUSIONS: Multivariate analysis showed that diabetes and residual stenosis (albeit <30%) were predictive of recurrence, whereas the finding of an isolated CAS lesion as opposed to stenoses in multiple locations was shown to be negatively associated with recurrent CAS, appearing to be "protective".