Prior reports on the COOK Zenith Alpha abdominal graft have shown limb graft occlusion (LGO) and stenosis (LGS) to be the primary causes of postoperative re-interventions. Special notices from the manufacturer have indicated certain IFU violations as the main cause of these complications. The current study aimed to assess the incidence of LGO and to analyze the impact of detailed anatomical risk factors for LGO highlighted in the instruction for use (IFU).

A retrospective single-center study was performed of all consecutive patients treated with the Zenith Alpha at a single institution from October 1, 2015, to September 30, 2018. All patients underwent computed tomography angiography (CTA) preoperatively as well as at three and 12 months postoperatively. Patients were then followed yearly with duplex ultrasound (DUS) and clinical exams. Additional CTAs were performed if dictated by clinical and/or DUS findings. All CTAs were analyzed using three-dimensional reconstruction software (TeraRecon.com). Data were extracted from electronic charts until the end of the study (December 31, 2020). LGS was defined as a significant stenosis resulting from a thrombus formation, or a narrowed limb graft lumen, assessed by the treating surgeon. LGO was defined as complete limb graft occlusion regardless of symptoms. The following outcomes were assessed: Rate of LGO, Kaplan Meier (KM) estimates of freedom from LGO, and LGO and LGS-related re-interventions. Cox regression analysis was performed to estimate the hazard ratio of risk factors associated with the development of LGO with 95% confidence intervals. P-values <0.05 were considered significant. The following risk factors were chosen based on IFUs and previously published reports: iliac tortuosity, iliac calcification, aortic diameter, iliac diameters, graft sizing, overlap of the graft components into main body, and distal sealing zone in the external iliac artery (EIA). The risk factors were defined according to current reporting standards. Due to the number of events, not all risk factors could be tested in the analysis simultaneously. Five risk factors were chosen for the core model. The remaining risk factors were then tested individually within the core model. Cohen's kappa was computed to evaluate reader agreement of the calcification of the common iliac arteries (CIAs) and the EIAs in patients with LGO.

241 patients were identified. 27 patients (11%) experienced LGO in 33 limbs (7%). The KM estimated freedom from LGO was 93% (CI 90-97), 90% (CI 86-94) and 88% (CI 84-92) at 1, 2 and 3 years postoperatively (Figure 1). Four patients presented with ischemic rest pain, 21 patients presented with intermittent claudication, and two patients were asymptomatic. 25 LGO-related re-interventions were performed in 20 patients (Table I). Seven patients were treated conservatively due to none to moderate symptoms (n=3), deemed surgically unfit (n=2), graft explantation due to rupture (n=1), and patient declining further intervention (n=1). In a cox regression analysis, calcified iliac CIAs and EIAs were negatively associated with LGO (Table II). The kappa values of the intra-reader agree¬¬¬ment (MB1 and MB2) of the CIA and the EIA calcification were 0.6 (95% CI: 0.5 - 0.8), p < .001, and 0.7 (95% CI: 0.5 - 0.9), p < 0.001, respectively. The kappa values of the inter-reader agree¬¬¬ment (MB1 and MT) of the CIA and the EIA calcification were 0.5 (95% CI: 0.3 - 0.7), p < 0.001, and 0.5 (95% CI: 0.4 - 0.7), p < 0.001, respectively.

The LGO-rate after EVAR with the Zenith Alpha graft was high. LGO development was not associated with any analyzed risk factors highlighted in the IFUs. The negative association of LGO to preoperative CIA calcification seems counterintuitive and might be explained by inadequate specificity in calcification measurements.