Type II endoleak after endoaortic graft implantation: diagnosis with helical CT arteriography.

Abstract

To retrospectively assess endoleak shapes and locations within aneurysms to differentiate type II from type I and type III endoleaks. The institutional review board granted an exemption for this HIPAA-compliant study; patient informed consent was not required. A retrospective review of arterial phase helical computed tomographic (CT) studies and medical records was performed for 39 patients (29 men, 10 women; age range, 60-89 years; mean, 78.5 years) who had an endoleak after endoaortic graft implantation for treatment of abdominal aortic aneurysm and who subsequently underwent angiography (n = 25), surgery (n = 8), or long-term follow-up (n = 6) to classify their endoleak into a specific type. At CT, endoleak shape (tubular or nontubular) and location (central or peripheral) were recorded. An endoleak was classified as type II if it contained a peripheral tubular component (PTC) near the aortic wall, with or without an identifiable feeding vessel. Endoleaks without these features were classified as type I or III. The Fisher exact test was used to assess associations between CT findings and endoleak type. There were 22 type II and 17 type I or III endoleaks. CT enabled correct identification of 22 (100%) of 22 type II endoleaks, all of which contained a PTC. Of 17 type I or III endoleaks, only two (12%) contained a PTC and were misclassified as type II endoleaks; the remaining 15 (88%) were correctly classified. Overall, CT enabled correct identification of endoleaks as type II or type I or III in 37 (95%) of 39 patients. PTCs were significantly more common (P < .001) in type II than in type I or III endoleaks, with a sensitivity, specificity, accuracy, negative predictive value, and positive predictive value of 100%, 88.2%, 94.9%, 100%, and 91.7%, respectively. A PTC is a statistically significant predictor of type II endoleak in most patients.