Factors Associated with Major Adverse Events after Fenestrated and Branched Endovascular Aortic Repair of Complex Abdominal and Thoraco-abdominal Aortic Aneurysms.

Abstract

Routine post-operative admission to the intensive care unit (ICU) is often advocated following fenestrated and/or branched endovascular aortic repair (FBEVAR) of complex abdominal aortic aneurysms (cAAAs) and thoraco-abdominal aortic aneurysms (TAAAs). Given limited ICU resources, identifying pre- and intra-operative predictors of major adverse events (MAE) is crucial for optimal resource allocation. Consecutive elective FBEVAR procedures performed between December 2012 and May 2020 were analysed retrospectively. Patients were divided into three groups based on aneurysm extent as cAAA, type 4 TAAA (TAAA4), and types 1 - 3 TAAA (TAAA1-3). MAEs were defined according to current Society for Vascular Surgery reporting standards. The primary endpoint was thirty day MAEs. Candidate pre- and intra-operative predictors were entered into a least absolute shrinkage and selection operator (LASSO) penalised logistic regression, followed by an unpenalised post-LASSO refit to report odds ratios (ORs). Four hundred and thirty-nine patients (129 cAAAs, 193 TAAA4, and 117 TAAA1-3) with 1 694 target arteries were included; 86.3% had four or more vessels incorporated. Primary technical success was 95.9% and the mean surgical time was 185 ± 68 minutes. MAEs occurred in 9.6% (n = 43), including a 3.6% mortality rate (n = 16). Most MAEs (75.0%) occurred within forty-eight hours. Grade 3 spinal cord ischaemia occurred in 1.6%, which was higher in patients with TAAA1-3 (4.3%; p = .023). The thirty day re-intervention rate was 8.2% (2.7% stent graft related; 3.4% access related). In the post-LASSO refit, women (OR 3.59, 95% confidence interval [CI] 1.07 - 10.8; p = .030), unplanned additional procedures (OR 2.62, 95% CI 1.18 - 5.75; p = .020), total fluoroscopy time (OR 1.67, 95% CI 1.20 - 2.29; p = .002), and norepinephrine use (OR 3.23, 95% CI 1.46 - 7.14; p = .003) were independently associated with MAEs. Model performance showed an area under the receiver operating characteristic curve of 0.78 (95% CI 0.69 - 0.85). This study suggests that women undergoing complex operations with unplanned additional procedures or extended total fluoroscopy time would benefit the most from close monitoring to detect MAEs in the immediate aftermath of FBEVAR. In contrast, men without these features might be candidates for standard monitoring.