Algorithm for the diagnosis and treatment of endoleaks

BACKGROUND: Endoluminal grafting of abdominal aortic aneurysms (AAA) has shown promising early results. However, endoleaks present a new and challenging obstacle to successful aneurysm exclusion. We report our experience with primary, persistent endoleaks and provide an algorithm for their diagnosis and management. METHODS: Over a 19-month period, 73 patients underwent endoluminal repair of their AAAs using a modular bifurcated endograft as part of a US FDA Investigational Device Exemption trial. Spiral computed tomography (CT) scanning was performed prior to discharge after repair to evaluate for complete aneurysm exclusion. If no endoleak was present on that initial CT scan, color-flow duplex scanning was performed at 1 month, with repeat CT scanning at 6 months and 1 year. If the initial CT scan revealed the presence of an endoleak, repeat CT scanning was performed at 2 weeks, 1 month, and 3 months, or until the endoleak resolved. Any patient with an endoleak that persisted beyond 3 months underwent angiographic evaluation to localize the source of the leak. RESULTS: At 1 month, 62 patients (85%) had successful aneurysm exclusion. The remaining 11 patients (15%) had primary endoleaks, 8 (11%) of which persisted beyond 3 months, prompting angiographic evaluation. In 2 patients the endoleak was related to a graft-graft or graft-arterial junction. One was from the endograft terminus in the common iliac artery and was successfully embolized, along with its outflow lumbar artery. The other required placement of an additional endograft component across a leaking graft-graft junction to successfully exclude the aneurysm. The remaining six endoleaks were due to collateral flow through the aneurysm sac. In 4 cases this was lumbar to lumbar flow fed by hypogastric artery collaterals to the inflow lumbar artery. In the remaining 2 patients the endoleak was found to be due to flow between a lumbar and inferior mesenteric artery. Resolution of the endoleak by coil embolization of the feeding hypogastric artery branch in 1 patient was unsuccessful due to rapid recruitment of another hypogastric branch. Two of the six collateral flow endoleaks have resolved spontaneously without treatment, while the remaining cases have been followed up without evidence of aneurysm expansion. CONCLUSION: Systematic postoperative surveillance facilitates proper diagnosis and treatment of endoleaks. This involves serial CT scans to detect the presence of endoleaks, followed by angiography to determine their etiology and guide treatment, if clinically indicated.