Occlusion of the Celiac Artery During Endovascular Thoracoabdominal Aortic Aneurysm Repair Is Associated with Increased Perioperative Morbidity and Mortality

Background: Some studies suggest that celiac artery coverage during elective endovascular thoracoabdominal aortic aneurysm (TAAA) repair is safe given sufficient collateralization of visceral organ perfusion from the superior mesenteric artery. However, there is concern that celiac artery coverage may lead to increased risk of foregut or spinal cord ischemia with an attendant increased risk of mortality. We sought to investigate rates of bowel ischemia, spinal cord ischemia, and 30-day mortality associated with celiac artery coverage during TEVAR and complex EVAR.

Methods: The Society for Vascular Surgery Vascular Quality Initiative database was queried for TEVAR and complex EVAR cases from 2012 to 2018. Inclusion criteria included TAAA pathology and endograft extension to aortic zone 6. Patients with aortic rupture, trauma, prior thoracic aortic surgery, known preoperative occlusion of the left subclavian superior mesenteric, or celiac arteries were excluded. Cases with intraoperative celiac artery occlusion (CAO) were compared retrospectively to cases with celiac artery preservation (CAP). Primary outcomes included 30-day mortality and a composite end point of 30-day mortality, spinal cord ischemia (transient or permanent lower extremity neurologic deficit), and bowel ischemia (colonoscopic evidence of ischemia, bloody stools in a patient who dies prior to colonoscopy or laparotomy, or other documented clinical diagnosis). Univariable comparisons were performed using chi-squared tests and Student's t-tests, as appropriate. Multivariable logistic regression analyses were employed to identify independent predictors of outcome.

Results: There were 628 cases identified for inclusion in the study. Patients undergoing CAO (n = 44) were more likely to be female or to have higher rates of preoperative spinal drain use, American Society of Anesthesiologists score ≥3, low preop hemoglobin, and/or symptomatic presentation, but fewer mean number of aortic zones covered. CAO was associated with higher 30-day mortality (5 of 44, 11%) compared to CAP (23 of 584, 4%), P = 0.039. The composite end point occurred at a significantly greater proportion for those who had CAO (10 of 44, 23%) compared to CAP (53 of 584, 9%, P = 0.008), driven by higher rates of 30-day mortality and bowel ischemia (9% vs. 2%, P = 0.026). By multivariate analysis, CAO was predictive of 30-day mortality (odds ratio [OR] = 3.9, 95% confidence interval [CI] = 1.1-13.8, P = 0.04) and the composite endpoint (OR = 3.0, 95% CI = 1.1-8.5, P = 0.03). Increasing procedure time was also associated with 30-day mortality (OR = 1.4, 95% CI = 1.1-1.7, P < 0.001) and the composite end point (OR = 1.4, 95% CI = 1.1-1.6, P < 0.001).

Conclusions: For those treated for TAAAs, CAO was independently predictive of increased 30-day mortality and a composite end point of perioperative mortality, spinal cord ischemia, and bowel ischemia. When treating patients with extensive aortic aneurysmal disease, physicians should attempt to preserve the celiac artery, by revascularization or avoiding ostium coverage, whenever feasible.