ndovascular abdominal aortic aneurysm repair (EVAR) has been established as a safe and effective

Vascular and Interventional Radiology Original Research Mursalin et al. Imaging-Based Predictors of Persistent Type II Endoleak After EVAR Vascular and Interventional Radiology Original Research Rafael Mursalin 1 Ichiro Sakamoto 1 Hiroki Nagayama 1 Eijun Sueyoshi 1 Kazuyoshi Tanigawa 2 Takashi Miura 2 Masataka Uetani 1 Mursalin R, Sakamoto I, Nagayama H, et al. Keywords: abdominal aortic aneurysm, CT, endoleak, endovascular abdominal aortic aneurysm repair, stent-graft DOI:10.2214/AJR.15.15254 Received July 11, 2015; accepted after revision December 9, 2015. 1 Department of Radiology, Graduate School of Biomedical Sciences, Nagasaki University, 1-7-1 Sakamoto, Nagasaki, 852-8501, Japan. Address correspondence to I. Sakamoto (ichiro-s@nagasaki-u.ac.jp). 2 Department of Cardiovascular Surgery, Graduate School of Medicine, Nagasaki University, Nagasaki, Japan. AJR 2016; 206:1335 1340 0361 803X/16/2066 1335 American Roentgen Ray Society Imaging-Based Predictors of Persistent Type II Endoleak After Endovascular Abdominal Aortic Aneurysm Repair OBJECTIVE. The purpose of this study is to determine the imaging-based parameters associated with the occurrence of persistent type II endoleaks after endovascular abdominal aortic aneurysm repair. MATERIALS AND METHODS. We reviewed the imaging and clinical data for 47 patients with early-onset type II endoleak after endovascular repair. Various predictors of persistent type II endoleaks were analyzed on the basis of preoperative CT findings. In addition, the appearance time of endoleak cavity on the operative angiogram and the relative attenuation of the endoleak cavity in the arterial phase image from the first postoperative CT study were analyzed. RESULTS. The early-onset type II endoleak resolved spontaneously in 22 patients (i.e., the transient group), whereas it was identified on CT studies of the remaining 25 patients 6 months after endovascular repair (i.e., the persistent group). The appearance time of the endoleak cavity on angiographic examination was significantly shorter in the persistent group than in the transient group (mean [± SD] appearance time, 4.7 ± 0.3 s vs 8.8 ± 0.3 s). The relative attenuation of the endoleak cavity on the first postoperative CT scan was also significantly higher in the persistent group than in the transient group (mean, 0.70 ± 0.03 vs 0.30 ± 0.04). For each parameter, ROC analysis revealed the following cutoff points for predicting persistent type II endoleak: 6 seconds for the appearance time of the endoleak cavity (sensitivity, 88%; specificity, 86%), and 0.5 for the relative attenuation of the endoleak cavity (sensitivity, 80%; specificity, 95%). Evaluation of other imaging-based parameters revealed no statistically significant differences between the groups. CONCLUSION. The appearance time of the endoleak cavity on the final operative angiogram and the attenuation of the endoleak cavity on the first postoperative CT scan can be strong imaging-based predictors of persistent type II endoleak. E ndovascular abdominal aortic aneurysm repair (EVAR) has been established as a safe and effective method of aneurysm repair because of its less invasive nature and its ability to significantly reduce the 30-day mortality rate, compared with open repair (1.7% vs 4.7%) [1, 2]. Despite the early positive outcome associated with the use of EVAR, abdominal aortic aneurysm (AAA) expansion may occur after EVAR because of a unique complication known as endoleak. Endoleak is defined as continued perfusion within the aneurysm sac despite the deployment of a stent-graft. On the basis of the origin of the leak, endoleak has been classified into five types [3, 4]. Of these different types of endoleaks, the type II endoleak is the most common and is caused by retrograde flow, which usually occurs in the lumbar and inferi- or mesenteric arteries [4 11]. Early-onset type II endoleaks can be classified as transient and persistent types [12]. By definition, transient type II endoleaks resolve spontaneously within 6 months of occurrence and are considered to be benign, and persistent type II endoleaks last more than 6 months [13]. Persistent type II endoleak may cause aneurysm sac enlargement, although not all patients with persistent type II endoleaks will necessarily have sac expansion. Such sac expansion has been reported to occur in 3%, 17%, and 41% of patients at 1, 3, and 5 years after EVAR, respectively [14]. In patients with aneurysm sac enlargement secondary to persistent type II endoleaks, aggressive techniques may be required to avoid eventual rupture of AAA; such techniques include percutaneous transarterial or direct translumbar embolization with microcoils or AJR:206, June 2016 1335

Mursalin et al. liquid embolic agents [15], endoscopic ligation of the lumbar or mesenteric arteries, and surgical explant (for severe cases) [16 18]. Hence, it is crucial to identify patients at high risk for persistent type II endoleak for whom vigilant surveillance may be recommended to avoid overlooking sac enlargement. Investigators have previously discussed various predictors of persistent type II endoleak, including the demographic and clinical characteristics of the patients (e.g., patient age and sex, the presence of cardiovascular disease, and the use of anticoagulant drugs, antiplatelet drugs, or both, and other characteristics) and imaging-based parameters (e.g., the size of the AAA, the volume of the intraaneurysmal thrombus, and the number and diameter of patent aortic side branches, and other parameters). However, to our knowledge, there currently is no reliable way to differentiate persistent type II endoleak from transient type II endoleak, because most previous studies did not identify predictors of persistent type II endoleak. In the past, most investigators discussed the use of imaging-based parameters for the prediction of persistent type II endoleak in terms of assessing preoperative CT findings only. There was little discussion of using findings from the operative angiogram and the first postoperative CT study as possible predictors. We postulated that different flow patterns (high flow vs low flow) in the endoleak cavity influence the outcomes of early-onset type II endoleaks, with those with a highflow pattern remaining present over a long period, and those with a low-flow pattern resolving spontaneously. Therefore, we evaluated the imaging features of early-onset type II endoleak on the operative angiogram and the first postoperative CT scan, in addition to the morphologic findings for AAA and the vascular anatomy of the patient, as observed on preoperative CT studies. The goal of the present study is to determine the imagingbased risk factors that are associated with persistent type II endoleak. Materials and Methods Patients This retrospective study, which was approved by the institutional review board at Nagasaki University, included 145 consecutive patients (115 men and 30 women; mean [± SD] age, 77.5 ± 1.9 years; range, 49 92 years) who underwent EVAR between October 2007 and May 2014. Data were obtained by searching the radiology information system and medical records at our institution. Of the 145 patients reviewed, early-onset type II endoleak was identified in 51 patients on the final operative angiogram. In all 51 patients, type II endoleak was also seen on the first postoperative CT scan, which was obtained 1 week after EVAR. All patients with less than 6 months of follow-up (n = 4) were excluded. Therefore, a total of 47 patients were included in the present study. All of the devices that were used were bifurcated, and two different stent-graft designs were used during the study period: the Excluder (W. L. Gore and Associates), which was used in 81 patients, and the Zenith (Cook Medical), which was used in 64 patients. A stent-graft was selected for use on the basis of the following principle: the Zenith, which possesses a suprarenal stent hook, was A preferred for patients with a short infrarenal neck, whereas the flexible Excluder was preferred for patients with a severely angulated infrarenal neck. However, device selection was ultimately determined according to physician preference. Imaging Studies Preoperative CT examinations included the area from the diaphragm to the femoral heads and were conducted using MDCT scanners (Somatom Flash Dual Source, Siemens Healthcare). The CT protocol involved the use of a triple-phase technique that included unenhanced, arterial, and delayed phases. Arterial phase images were generated during injection of 100 ml of nonionic contrast material at a flow rate of 4 ml/s, with the use of bolus tracking with C D Fig. 1 72-year-old man with persistent type II endoleak after endovascular abdominal aortic aneurysm repair (EVAR). A and B, Final operative angiograms obtained at 2 seconds (A) and 7 seconds (B) after injection of contrast material into abdominal aorta show type II endoleak (arrow, B) from bilateral fourth lumbar arteries. In this case, time interval between appearance of renal artery and type II endoleak was relatively short (5 seconds), thus suggesting high-flow pattern in endoleak cavity. C, Early phase contrast-enhanced CT image obtained 1 week after EVAR shows type II endoleak (arrow) with dense accumulation of contrast material indicating high-flow pattern. D, Early phase contrast-enhanced CT obtained 6 months after EVAR shows type II endoleak cavity (arrow), which increased in size during follow-up. B 1336 AJR:206, June 2016

Imaging-Based Predictors of Persistent Type II Endoleak After EVAR TABLE 1: Demographic Characteristics of 47 Patients Evaluated Characteristic Patients With Persistent Endoleak (n = 25) an attenuation threshold of 150 HU. Delayed phase images were obtained 90 seconds after the arterial phase scan was acquired. For image analysis, transverse and coronal plane images were reconstructed with a section thickness of 3.0 mm and a section increment of 3.0 mm. The first postoperative CT scan was obtained 1 week after EVAR. Follow-up CT was commonly performed at 3, 6, and 12 months after EVAR and at 6-month intervals thereafter. All CT scans obtained during follow-up were triple-phase scans, which were acquired as described previously. Evaluation of Imaging-Based Risk Factors Measurements of imaging-based risk factors for persistent type II endoleaks were performed by two experienced radiologists who reached consensus by using a separate graphical workstation (syngo.via, Siemens Healthcare) from the CT scanner. The readers were blinded to the outcomes (identification of transient vs persistent type) of early-onset type II endoleaks. The following risk factors were analyzed on the basis of preoperative CT images: the size of the AAA; the number and patency of the inferior mesenteric arteries, the internal iliac arteries, and the lumbar arteries; and the maximum diameter of the largest lumbar artery. The size of the AAA and the diameter of the largest lumbar artery were measured on axial sections with the use of electronic calipers. In addition, the maximum area of type II endoleak was measured in the delayed phase of the first postoperative CT study. A line was drawn manually around the endoleak cavity to define the cavity outline on each slice, and the maximum area of the endoleak cavity was determined. The appearance time of the endoleak cavity on the final operative angiogram and the relative attenuation of the endoleak cavity on the arterial phase image from the first postoperative CT study were measured because these parameters were considered to reflect flow velocity in the endoleak cavity. The appearance time of the endoleak cavity was determined on the basis of the interval between visualization of the renal artery and the endoleak cavity on the final operative angiogram. The relative attenuation of the endoleak cavity on the arterial phase image from the first postoperative CT study Patients With Transient Endoleak (n = 22) Age (y), mean ± SD 75.8 ± 1.7 79.4 ± 1.8 0.1210 Sex, no. of patients 0.6192 Male 18 18 Female 7 4 a Statistically significant. was calculated on the basis of the following formula: (endoleak cavity attenuation on the early phase image endoleak cavity attenuation on the unenhanced image) / (attenuation of the stent-graft lumen on the early phase image attenuation of the stent-graft lumen on the unenhanced image), where attenuation is measured in Hounsfield units. The outcome (identification of transient vs persistent type) of early-onset type II endoleak was determined on the basis of findings from a contrast-enhanced CT study obtained 6 months after EVAR. The imaging-based risk factors described previously in the present study were analyzed to determine which factors were associated with persistent type II endoleak. Statistical Analysis Statistical analysis was conducted using statistical software (JMP 12.0, SAS Institute). Values were expressed as mean values and SDs or as percentages, as appropriate. Univariate and multivariate analyses were performed using logistic regression analysis to identify the factors associated with persistent type II endoleak. Only variables for which p < 0.1 on univariate analysis were entered into the multivariate logistic regression model. A p value of 0.05 or lower was considered to denote statistical significance. To assess the diagnostic accuracy of features noted on the final operative angiogram and the first postoperative CT scan for predicting persistent type II endoleak, ROC plots were analyzed and AUCs were calculated. The cutoff points of these parameters for predicting persistent type II endoleak were calculated using ROC analysis. p a Results For 47 patients, the early-onset type II endoleak was identified on the final operative angiogram, and for all 51 patients, it was detected on the first postoperative CT scan. In 22 of the 45 patients, the early-onset type II endoleak resolved spontaneously; however, in the remaining 25 patients, it was identified on a CT scan obtained 6 months after EVAR. Therefore, a total of 22 patients had a transient type II endoleak and were known as the transient group, and 25 patients had a persistent type II endoleak and were known as the persistent group. The characteristics of each group at baseline are shown in Table 1. No statistically significant differences in sex and age were noted between the transient and persistent groups. The early-onset type II endoleak was identified in 32 of 81 patients (40%) who received the Excluder device. Of these 32 patients, 19 had persistent type II endoleaks and 13 had transient type II endoleaks. The early-onset type II endoleak was identified in 15 of 64 patients (23%) who received the Zenith device, six of whom had a persistent endoleak and nine of whom had a transient endoleak. Hence, the frequency of early-onset type II endoleaks was statistically significantly greater among patients who received the Excluder device (40%) than among patients who received the Zenith device (23%) (p < 0.05). In addition, the frequency of persistent type II endoleak was also statistically significantly greater among patients who received the Excluder device (19/81 [23%]) than among patients who received the Zenith device (6/64 [9%]) (p < 0.05). The relationship between the stent-graft devices used and the type II endoleak identified is shown in Table 2. Findings from univariate and multivariate analyses of preoperative CT images, the final operative angiogram, and the first postoperative CT images are presented in Table 3. Univariate analysis revealed that the diam- TABLE 2: Relationship Between Stent-Graft Devices Used and Development of Early-Onset Type II Endoleaks Variable Excluder Device (n = 81) Zenith Device (n = 64) p a Absence of type II endoleaks 49 (60) 49 (77) Early-onset type II endoleak Transient type 13 (17) 9 (14) Persistent type 19 (23) 6 (9) < 0.05 All 32 (40) 15 (23) < 0.05 Note Data are number (%) of patients with endoleaks. a Statistically significant. AJR:206, June 2016 1337

Mursalin et al. TABLE 3: Results of Univariate and Multivariate Analysis of Imaging-Based Predictors for Persistent Type II Endoleak p eter of the largest lumbar artery was significantly larger in the persistent group than in the transient group (mean, 2.4 ± 0.1 vs 2.1 ± 0.1 mm). However, regarding this parameter, multivariate analysis found no statistically significant difference between the transient and persistent groups. The appearance time of the endoleak cavity on the final operative angiogram was significantly shorter in the persistent group than in the transient group on both univariate and multivariate analysis (mean, 4.7 ± 0.3 vs 8.8 ± 0.3 s). The relative attenuation of the endoleak cavity in the arterial phase image from the first postoperative CT study was also higher in the persistent group than in the transient group on both univariate and multivariate analysis (mean, 0.70 ± 0.03 vs 0.30 ± 0.04 HU) (Fig. 1). On the other hand, preoperative CT features, such as the maximum diameter of the AAA, presence of a patent inferior mesenteric artery, the total number of patent lumbar arteries, and the total number of patent internal iliac arteries, were not statistically significantly different between the persistent and transient groups. The endoleak cavity area in the delayed phase image from the first postoperative CT study was not statistically significantly different between the persistent and transient groups on both univariate and multivariate analysis. The ROC curves show the fraction of truepositive results (sensitivity) and false-positive results (1 specificity) for various cutoff levels of the appearance time of the endoleak cavity on the final operative angiogram and the relative attenuation of the endoleak cavity on the arterial phase image from the first Variable Persistent Endoleak Transient Endoleak postoperative CT study, for prediction of persistent type II endoleak (Figs. 2 and 3). For each parameter, ROC analysis revealed the following cutoff points for predicting persistent type II endoleak: 6 seconds, for the appearance time of the endoleak cavity on the final operative angiogram (sensitivity, 88%; specificity, 86%), and 0.5 HU, for the relative attenuation of the endoleak cavity in the arterial phase of the first postoperative CT scan (sensitivity, 80%; specificity, 95%). Discussion The appearance time of the type II endoleak on the final operative angiogram (i.e., the interval between visualization of the renal artery and the endoleak cavity) and the relative attenuation of the endoleak cavity in the arterial phase of the first postoperative CT image were used as indicators of flow patterns in endoleak cavity, and these parameters were statistically significantly associated with outcomes of persistent type II endoleak. In addition, ROC analysis also revealed the following cutoff points for predicting persistent type II endoleak: 0.5, for relative attenuation of endoleak cavity at arterial phase of first postoperative CT (sensitivity, 80%; specificity, 95%), and 6 seconds, for the interval between visualization of the renal artery and the endoleak cavity (sensitivity, 88%; specificity, 86%). Accordingly, these two parameters seem to be useful as predictors for persistent type II endoleak. The number of patent aortic side branches had no predictive value for persistent type II endoleak in the present study, a result that was not consistent with findings from a 2001 Univariate Analysis Multivariate Analysis Abdominal aortic aneurysm size (mm) 52.8 ± 1.8 51.0 ± 1.9 0.4996 No. of patent lumbar arteries 6.4 ± 0.3 6.1 ± 0.4 0.5683 Maximum diameter of the lumbar arteries (mm) 2.4 ± 0.1 2.1 ± 0.1 0.05 a 0.1121 Patency status of inferior mesenteric artery (no. of arteries) 0.2864 Occluded 5 2 Patent 20 20 Area of endoleak (mm 2 ) 86.3 ± 73.0 57.0 ± 64.8 0.1366 No. of patent internal iliac arteries 1.5 ± 0.14 1.6 ± 0.15 0.5543 Relative attenuation of endoleak cavity on first postoperative CT (HU) 0.70 ± 0.03 0.30 ± 0.04 < 0.0001 a 0.0186 a Appearance time of endoleak cavity on final operative angiogram (s) 4.7 ± 0.3 8.8 ± 0.3 < 0.0001 a 0.0039 a Note Except where indicated otherwise, data are mean ± SD values. Statistically significant. study by Fan et al. [19], who reported that development of early-onset type II endoleak was statistically significantly associated with the number of the aortic side branches. However, Fan and colleagues did not examine the outcomes (transient vs persistent type) of the early-onset type II endoleak. Therefore, patent side branches can be a predictor of the development of early-onset type II endoleak, although they may be of little value in predicting the outcome of early-onset type II endoleak. Keedy et al. [20] suggested that the diameter of the largest communicating vessels was significantly associated with the clinical outcome of early-onset type II endoleak, although that relationship was statistically weak. In the present study, the diameter of the largest lumbar arteries was larger in the persistent group than in the transient group on univariate analysis, although it was not statistically significantly associated with persistent type II endoleak on multivariate analysis. In our experience, measuring the diameters of all side branches with the use of digital calipers is time consuming and seems to be affected by intrinsic inaccuracies. Therefore, this parameter may be difficult to translate easily to the clinical setting. The Excluder device was statistically significantly associated with a higher rate of early-onset type II endoleaks, compared with the Zenith device. In addition, this device was associated with a statistically significantly higher rate of persistent type II endoleaks. These findings were consistent with those of Ouriel et al. [21], who suggested that the Excluder device was associated with a higher rate of type II endoleaks and a decreased rate of aneurysm 1338 AJR:206, June 2016

Imaging-Based Predictors of Persistent Type II Endoleak After EVAR 100 90 80 100 90 80 Sensitivity (%) 70 60 50 40 30 20 10 0 regression, compared with other endovascular grafts. However, we must consider the fact that a particular device was selected on the basis of specific anatomic configurations, potentially creating a biased outcome for or against that device when interpreting the result of device-specific differences. The results showed that the maximum area of the endoleak cavity on the first postoperative CT scan was not statistically significantly associated with a higher likelihood of persistent type II endoleak. Dudeck et al. [22] suggested that the size of the endoleak volume was of prognostic relevance, whereas the simple assessment of diameter and area on the section of the greatest endoleak extension was not of predictive value. Certainly, the size of the endoleak volume may be a useful predictor for persistent type II endoleaks. However, assessment of the endoleak volume seems to be impractical in routine clinical practice because it is time-consuming and demanding. The goal of the present study, therefore, is to look for a practical surrogate predictor of persistent type II endoleaks. The present study shows that the appearance time of the endoleak cavity on final operative angiography and attenuation of the endoleak cavity on the arterial phase image from the first postoperative CT study can be an indicator for persistent type II endoleak. These parameters, especially the appearance Sensitivity = 88% Specificity = 86% AUC = 0.956 0 20 40 60 80 100 1 Specificity (%) Fig. 2 ROC analysis was used to determine optimal cutoff point for appearance time of endoleak cavity on final operative angiogram, for prediction of persistent type II endoleak. Point estimates at curve indicate cutoff values (sensitivity and specificity). Sensitivity (%) time of the endoleak cavity on the final operative angiogram, can readily be assessed and seem less likely to be affected by observer variability, compared with previously reported imaging-based parameters. Therefore, these parameters can be practically used as predictors for persistent type II endoleaks and may help differentiate patients who need close monitoring from those who require infrequent follow-up evaluation. Furthermore, research has suggested that patients for whom the presence of endoleak has been excluded on review of the first postprocedure three-phase CT images (an unenhanced image plus images from both contrastenhanced phases) can undergo unenhanced CT examinations only for further surveillance [23]. This approach offers the benefits of reduced cost, avoidance of the risk of exposure to iodinated contrast material, and decreasing radiation exposure for patients with stable AAA after EVAR. The results of the present study uggest that patients with early-onset type II endoleak can also be monitored by the use of unenhanced CT examinations only if a final operative angiogram or first postoperative CT scan indicates a low-flow pattern in endoleak cavity. However, large prospective studies are needed to substantiate this contention, because the sample size of our study was rather small. Finally, other studies [24, 25] have encouraged the use of prophylactic sac embolization 70 60 50 40 30 20 10 0 Sensitivity = 80% Specificity = 95% AUC = 0.944 0 20 40 60 80 100 1 Specificity (%) Fig. 3 ROC analysis was used to determine optimal cutoff point for attenuation of endoleak cavity on first postoperative CT scan, for prediction of persistent type II endoleak. Point estimates at curve indicate cutoff values (sensitivity and specificity). during EVAR for the prevention of type II endoleaks. However, sac embolization is not widely performed, because this technique is technically time consuming and is associated with a certain risk of complications. If this technique is applied to selected patients with a relatively rapid appearance of endoleak cavity on a final operative angiogram, unnecessary treatment may be avoidable. The present study has some limitations. First, the major limitation of the study is the retrospective design of the analysis. Second, early-onset type II endoleak only was included; thus, our results cannot be applied to patients with late-onset type II endoleak. Third, endoleak cavity attenuation was calculated from findings from the first operative CT scan obtained 1 week after EVAR. On the other hand, at many institutions, the first postoperative CT scan is commonly obtained 1 month after EVAR. The 1-week follow-up CT seems to be preferable to the 1-month follow-up CT, because the 1-week follow-up CT enables early decision making regarding treatment of patients after EVAR. However, further studies are needed to substantiate the importance of the 1-week follow-up CT, compared with the 1-month follow-up CT. Fourth, attenuation adjacent to the vertebral body may be influenced by beam-hardening artifacts, thus causing inaccurate attenuation measurements in the endoleak cavity. In the present study, beam- AJR:206, June 2016 1339

Mursalin et al. hardening artifacts were relatively minimal. Hence, for all patients, obtaining measurements of attenuation in the endoleak cavity was feasible with the use of a selected axial image that was less affected by beam-hardening artifacts. Finally, analysis of imagingbased risk factors for persistent type II endoleak was done by consensus review without documentation of observer variability. In conclusion, the appearance time of the endoleak cavity on the final operative angiogram and the attenuation of the endoleak cavity in the arterial phase of the first postoperative CT scan can be strong imagingbased parameters that can be used to indicate which early-onset type II endoleaks will resolve spontaneously and which will last longer than 6 months. References 1. Greenhalgh RM, Brown LC, Kwong GP, et al. Comparison of endovascular aneurysm repair with open repair in patients with abdominal aortic aneurysm (EVAR trial 1), 30-day operative mortality results: randomised controlled trial. Lancet 2004; 364:843 848 2. United Kingdom EVAR Trial Investigators. Endovascular versus open repair of abdominal aortic aneurysm. N Engl J Med 2010; 362:1863 1871 3. White GH, May J, Waugh RC, Yu W. Type I and type II endoleaks: a more useful classification for reporting results of endoluminal AAA repair. J Endovasc Surg 1998; 5:189 191 4. White GH, Yu W, May J. Endoleak: a proposed new terminology to describe incomplete aneurysm by an endoluminal graft. J Endovasc Surg 1996; 3:124 125 5. Greenberg RK, Chuter TA, Sternbergh WC III, Fearnot NE; Zenith Investigators. Zenith AAA endovascular graft: intermediate term results of the US multicenter trial. J Vasc Surg 2004; 39:1209 1218 6. Wahlgren CM, Malmstedt J; Swedish Vascular Registry. Outcome of endovascular aortic aneurysm repair compared with open surgical repair in high-risk patients: results from the Swedish Vascular Registry. J Vasc Surg 2008; 48:1382 1388 7. Madden N, Baril DT, Wertz R, DeMarsico A, Rhee RY, Chaer RA. Endovascular abdominal aortic aneurysm repair: a community hospital s experience. Vasc Endovascular Surg 2009; 43:25 29 8. White GH, Yu W, May J, Chaufour X, Stephen MS. Endoleak as a complication of endoluminal grafting of abdominal aortic aneurysms: classification, incidence, diagnosis and management. J Endovasc Surg 1997; 4:152 168 9. Warrier R, Miller R, Bond R, Robertson IK, Hewitt P, Scott A. Risk factors for type II endoleaks after endovascular repair of abdominal aortic aneurysms. ANZ J Surg 2008; 78:61 63 10. Kranokpiraksa P, Kaufman JA. Follow-up of endovascular aneurysm repair: plain radiography, ultrasound, CT/CT angiography, MR imaging/ MR angiography, or what? J Vasc Interv Radiol 2008; 19(suppl 6):S27 S36 11. Stavropoulos SW, Charagundla SR. Imaging techniques for detection and management of endoleaks after endovascular aortic aneurysm repair. Radiology 2007; 243:641 655 12. Jones JE, Atkins MD, Brewster DC, et al. Persistent type 2 endoleak after endovascular repair of abdominal aortic aneurysm is associated with adverse late outcomes. J Vasc Surg 2007; 46:1 8 13. Bade MA, Ohki T, Cynamon J, Veith FJ. Hypogastric artery aneurysm rupture after endovascular graft exclusion with shrinkage of the aneurysm: significance of endotension from a virtual, or thrombosed type II endoleak. J Vasc Surg 2001; 33:1271 1274 14. Schanzer A, Greenberg RK, Hevelone N, et al. Predictors of abdominal aortic aneurysm sac enlargement after endovascular repair. Circulation 2011; 123:2848 2855 [Erratum in Circulation 2012; 125:e266] 15. Baum RA, Carpenter JP, Tuite CM, et al. Diagnosis and treatment of inferior mesenteric arterial endoleaks after endovascular repair of abdominal aortic aneurysms. Radiology 2000; 215:409 413 16. van Marrewijk CJ, Fransen G, Laheij RJ, Harris PL, Buth J; EUROSTARS Collaborators. Is a type II endoleak after EVAR a harbinger of risk? Causes and outcome of open conversion and aneurysm rupture during follow-up. Eur J Vasc Endovasc Surg 2004; 27:128 137 17. Faries PL, Cadot H, Agarwal G, Kent KC, Hollier LH, Marin ML. Management of endoleak after endovascular aneurysm repair: cuffs, coils and conversion. J Vasc Surg 2003; 37:1155 1161 18. Terramani TT, Chaikof EL, Rayan SS, et al. Secondary conversion due to failed endovascular abdominal aortic aneurysm repair. J Vasc Surg 2003; 38:473 477; discussion, 477 478 19. Fan CM, Rafferty EA, Geller SC, et al. Endovascular stent-graft in abdominal aortic aneurysm: the relationship between present vessels that arise from the aneurysm sac and early endoleak. Radiology 2001; 218:176 182 20. Keedy AW, Yeh BM, Kohr JR, Hiramoto JS, Schneider DB, Breiman RS. Evaluation of potential outcome predictors in type II endoleak: a retrospective study with CT angiography feature analysis. AJR 2011; 197:234 240 21. Ouriel K, Clair DG, Greenberg RK, et al. Endovascular repair of abdominal aortic aneurysms: devicespecific outcome. J Vasc Surg 2003; 37:991 998 22. Dudeck O, Schnapauff D, Herzog L, et al. Can early computed tomography angiography after endovascular aortic aneurysm repair predict the need for reintervention in patients with type II endoleak? Cardiovasc Intervent Radiol 2015; 38:45 52 23. Bley TA, Chase PJ, Reeder SB, et al. Endovascular abdominal aortic aneurysm repair: nonenhanced volumetric CT for follow-up. Radiology 2009; 253:253 262 24. Ronsivalle S, Faresin F, Franz F, Rettore C, Zanchetta M, Olivieri A. Aneurysm sac thrombization and stabilization in EVAR: a technique to reduce the risk of type II endoleak. J Endovasc Ther 2010; 17:517 524 25. Piazza M, Frigatti P, Scrivere P, et al. Role of aneurysm sac embolization during endovascular aneurysm repair in the prevention of type II endoleak-related complications. J Vasc Surg 2013; 57:934 941 1340 AJR:206, June 2016