Analysis of predisposing factors for type III endoleaks from directional branches after branched endovascular repair for thoracoabdominal aortic aneurysms

Background: Mid-term durability of branches has already been established, and BF-branched and fenestrated endovascular repair has shown comparable results with open repair in the treatment of thoracoabdominal aortic aneurysms (TAAAs). Nevertheless, target vessel instability remains the most frequent adverse event after complex endovascular aortic repair. Type III endoleaks from directional branches have been reported with a low incidence, but risk factors for this complication have not been investigated yet. Methods: This was a dual-center observational retrospective cohort study. Data were collected prospectively for each patient treated with branched endovascular repair between April 2008 and December 2019. The primary outcome was to assess potential risk factors for branch disconnection and fracture. A logistic regression analysis was performed, including preoperative and postoperative measurements as well as intraoperative details. A Cox regression hazard analysis was performed to evaluate the influence of preoperative aneurysm diameter and target vessel angulation on the outcome during follow-up. Results: Two hundred ninety-five target visceral vessels (TVVs) in 91 patients were considered suitable for cannulation. Technical success was 96.9% (286/295 TVVs). The median follow-up was 32.5 months (interquartile range, 14.2-50.1 months). Twelve type III endoleaks from directional branches were detected (4.2%; 5 bridging stent graft fractures and 7 disconnections). Five type III endoleaks involved the celiac trunk (one fracture and four disconnections), five the superior mesenteric artery (four fractures and one disconnection), and two the renal arteries (both disconnections). The median time to type III endoleak was 22.2 months (interquartile range, 10.9-37.6 months). Preoperative TAAA diameter (P = .028), preoperative TVV angulation (P = .037), the use of a BeGraft stent graft as bridging stent graft (P = .001), and different stent types on the same vessel (P = .048) were associated with type III endoleak at univariable analysis. Using a BeGraft stent graft (P = .010) was the only significant factor predisposing to type III endoleak at multiple logistic regression. The Cox regression analysis showed a two-fold increased risk for type III endoleak for every 10-mm increase in preoperative TAAA diameter (hazard ratio, 2.00; 95% confidence interval, 1.08-3.72; P = .028) and a 1.5 increased risk every 12° increase of preoperative TVV angulation (hazard ratio, 1.47; 95% confidence interval, 1.02-2.10; P = .037). Conclusions: Type III endoleaks from directional branches are a non-negligible complication after branched endovascular repair, with a relevant incidence. They tended to be clustered on specific patients, and aneurysm diameter and TVV angulation are strictly associated with the outcome. Different stent types on the same vessel should be avoided whenever possible. An intensified follow-up should be adopted for patients with large aneurysms, implanted with first-generation BeGraft, or who have been already diagnosed with type III endoleaks.