Occult type I or III endoleaks are a common cause of failure of type II endoleak treatment after endovascular aortic repair

Objective: Most type II endoleaks have a benign natural history, but 6% to 8% are associated with sac enlargement and respond poorly to treatment. Our aim was to evaluate whether these enlargements are associated with delayed or occult type I and III endoleaks. Methods: Patients with interventions for endoleak after endovascular aortic repair from 2000 to 2016 were reviewed retrospectively. Patient demographics, comorbidities, endoleak type, secondary procedures, aortic sac growth (>= 5 mm), and mortality were collected. Successful treatment was defined as endoleak resolution with no further aortic sac growth. Secondary procedures, ruptures, endograft explant, and death were captured. Results: There were 130 patients diagnosed with a primary type II endoleak after endovascular aortic repair at a median of 1.3 months (interquartile range, 1.0-13.3 months). One hundred eighteen had their initial treatment for a primary type II. Twelve of the 130 were initially stable and observed, but were treated for a delayed type I or III endoleak. The 130 patients underwent 279 procedures for endoleaks (mean of 2.2 +/- 1.3) over 6.9 +/- 3.8 years of follow-up. Of the 118 patients treated for primary type II endoleaks, 26 (22.0%) later required interventions for delayed type I and III endoleaks. The mean time to intervention for a delayed type I or III endoleak was 5.4 +/- 2.8 years. Overall, there were 16 type IA, 11 type IB, 2 type III, 7 combined type IA/IB, and 2 type IA/III delayed endoleaks. The odds of harboring a delayed type I or III endoleak was 22.0% before the first attempt at type II endoleak treatment, 35.1% before the second, 44.8% before the third, and 66.6% before the fourth attempts. Rapid aortic sac growth of >= 5 mm/y before initial endoleak treatment was associated with increased risk for delayed type I or III endoleak (47.8 vs 14.1%; P =.003). Patients with delayed type I or III endoleaks had a lower successful treatment rate (8.3% vs 52.3%; P =.001) than those with only type II endoleaks. Late rupture was increased with delayed type I or III endoleak (P =.002), whereas mortality (P =.96) and aortic-related mortality (P =.46) were similar. Graft explant (P =.06) trended toward an increase with a delayed type I or III endoleak, but was not statistically significant. Conclusions: Failed attempts treating type II endoleaks and/or a rapid aortic sac growth of 5mm/y or greater should raise the suspicion of a delayed or occult type I or III endoleak. Occult endoleaks are associated with decreased chance of endoleak resolution.