- Email: [email protected]
Predicting aneurysm enlargement in patients with persistent type II endoleaks
From the New England Society for Vascular Surgery
Predicting aneurysm enlargement in patients with persistent type II endoleaks Carlos H. Timaran, MD, Takao Ohki, MD, PhD, Soo J. Rhee, MD, Frank J. Veith, MD, Nicholas J. Gargiulo III, MD, Hisako Toriumi, MD, Mahmood B. Malas, MD, William D. Suggs, MD, Reese A. Wain, MD, and Evan C. Lipsitz, MD, New York, NY Objective: The clinical significance of type II endoleaks is not well understood. Some evidence, however, indicates that some type II endoleaks might result in aneurysm enlargement and rupture. To identify factors that might contribute to aneurysm expansion, we analyzed the influence of several variables on aneurysm growth in patients with persistent type II endoleaks after endovascular aortic aneurysm repair (EVAR). Methods: In a series of 348 EVARs performed during a 10-year period, 32 patients (9.2%) developed type II endoleaks that persisted for more than 6 months. Variables analyzed included those defined by the reporting standards for EVAR (SVS/AAVS) as well as other endoleak characteristics. Univariate, receiver operating characteristic curve, and Cox regression analyses were used to determine the association between variables and aneurysm enlargement. Results: The median follow-up period was 26.5 months (range, 6-88 months). Thirteen patients (41%) had aneurysm enlargement by 5 mm or more (median increase in diameter, 10 mm), whereas 19 (59%) had stable or shrinking aneurysm diameter. Univariate and Cox regression analyses identified the maximum diameter of the endoleak cavity, ie, the size of the nidus as defined on contrast computed tomography scan, as a significant predictor for aneurysm enlargement (relative risk, 1.12; 95% confidence interval, 1.04-1.19; P ⴝ .001). The median size of the nidus was 23 mm (range, 13-40 mm) in patients with aneurysm enlargement and 8 mm (range, 5-25 mm) in those without expansion (Mann-Whitney U test, P < .001). Moreover, receiver operating characteristic curve and Cox regression analyses showed that a maximum nidus diameter greater than 15 mm was particularly associated with an increased risk of aneurysm enlargement (relative risk, 11.1; 95% confidence interval, 1.4-85.8; P ⴝ .02). Other risk factors including gender, smoking history, hypertension, need of anticoagulation, aneurysm diameter, type of endograft used, and number or type of collateral vessels were not significant predictors of aneurysm enlargement. Conclusions: In patients with persistent type II endoleaks after EVAR, the maximum diameter of the endoleak cavity or nidus is an important predictor of aneurysm growth and might indicate the need for more aggressive surveillance as well as earlier treatment. (J Vasc Surg 2004;39:1157-62.)
The most frequent mechanism of failure after endovascular aortic aneurysm repair (EVAR) is the occurrence of an endoleak, ie, the persistence of blood flow in the aneurysm sac outside the endograft.1,2 Type II endoleaks are attributed to retrograde flow from patent lumbar arteries, inferior mesenteric artery, and other collateral aortic branches.2-4 Although type II endoleaks are frequently evident in intraoperative angiograms after EVAR, the majority of these endoleaks resolve as a result of the reversal of anticoagulation and the subsequent thrombosis of the aneurysm sac and its side branches.2 Persistent type II endoleaks develop in approximately 5% to 25% of the patients after EVAR.1,2,5-9 The exact clinical significance of persistent type II endoleaks is not defined because these might be associated with aneurysm enlargement, stability, and From the Division of Vascular Surgery, Department of Surgery, Montefiore Medical Center, Albert Einstein College of Medicine. Competition of interest: none. Presented at the Thirtieth Anniversary Meeting of the New England Society for Vascular Surgery, Newport, RI, Sep 19-21, 2003. Winning paper of the Darling Award of the New England Society for Vascular Surgery. Reprint requests: Takao Ohki, MD, Montefiore Medical Center, 111E 210th St, Bronx, NY 10467 (e-mail: [email protected]). 0741-5214/$30.00 Copyright © 2004 by The Society for Vascular Surgery. doi:10.1016/j.jvs.2003.12.033
shrinkage.1,2,8,10 Therefore, the optimal method of diagnosis and treatment of type II endoleaks is also controversial.9 As more experience with the diagnosis and treatment of type II endoleaks has been reported, it is now well recognized that type II endoleaks are not always benign. For instance, recent studies have documented that type II endoleaks might produce systemic blood pressures within the aneurysm sac that might result in aneurysm enlargement and rupture.4,8,11-14 Because the risk of aneurysm rupture is related to the maximum diameter of the aneurysm, general agreement exists that persistent type II endoleaks associated with aneurysm enlargement require aggressive management. However, it is still unknown which type II endoleaks will result in aneurysm expansion, thereby increasing the risk of rupture. This study was designed to assess the influence of several variables on aneurysm expansion in patients with persistent type II endoleaks after EVAR to define when a more aggressive approach of this type of endoleak is indicated. The most current recommended reporting standards were used to define the different variables.15,16 METHODS Among 348 patients who underwent EVAR during a 10-year period (November 1992 to September 2002), 32 1157
1158 Timaran et al
JOURNAL OF VASCULAR SURGERY June 2004
Fig 1. Maximum diameter measurements of the main endoleak cavity or nidus (e) was obtained from CT scan images by using calipers or CT imaging analysis software with magnification and precise cursor placement. A, Small endoleak nidus (2) measuring 9 mm. Although the endoleak channel was sometimes irregular and complex, a maximum nidus diameter could be determined in all patients. B, Large endoleak nidus measuring 23 mm in maximum diameter. When more than one nidus was seen, the maximum diameter of the largest nidus was selected.
patients (9.2%) developed type II endoleaks that were initially observed at the first postoperative month or at any time thereafter and persisted for more than 6 months. These 32 patients with persistent type II endoleaks were included in a cohort study. Patients with type II endoleaks associated with either type I or type III endoleaks at any time were specifically excluded. Patient clinical and procedural characteristics and follow-up results of patients undergoing EVAR were prospectively collected. The medical records, computed tomography (CT) scans, and angiographic studies of these patients were, however, further reviewed to define clinical characteristics, risk factors, morphologic features of endoleaks, and changes in aortic aneurysms according to the current reporting standards for EVAR prepared and revised by the Ad Hoc Committee for Standardized Reporting Practices in Vascular Surgery (SVS/AAVS).15,16 Patients with persistent type II endoleaks underwent CT scans every 3 to 6 months, and no patients were lost to follow-up. Axial thin-slice (1.5-3 mm thick) CT imaging and occasionally three-dimensional (3D) CT reconstruction models were obtained to confirm the presence of an endoleak and to further define the etiology and the most likely classification of the leak. Sequential changes in aneurysm size were determined with serial measurements of the maximum aneurysm diameter by using a standardized method previously described to decrease interobserver variability.17 Patients with persistent type II endoleaks who had documented aneurysm enlargement on follow-up CT scanning underwent diagnostic
angiography to characterize the endoleak. Once a type II endoleak was confirmed angiographically in patients with evidence of aneurysm expansion, additional treatment with either transarterial or translumbar embolization was performed. The technique of these treatment modalities has been described in detail before. Follow-up dynamic CT scans with 3-mm slice-thickness were obtained at 3, 6, and 12 months and annually thereafter. Embolization was considered to be technically successful when resolution of the endoleak was obtained with stabilization or shrinkage of the aneurysm sac. All follow-up CT scans and angiograms of patients with persistent type II endoleaks were reviewed to characterize the endoleak channel in greater detail. Endoleak angiography and pressure measurements were available in patients who underwent translumbar embolization. The number, origin, and outflow vessels feeding the endoleak were specified as suggested by the current reporting standards.15 The size of the main endoleak cavity or nidus was recorded including the maximum diameter, short axis, and length of the endoleak cavity. Calipers or CT imaging analysis software with magnification and precise cursor placement were used for diameter measurements obtained from CT scan images (Fig 1). Descriptive statistics for categorical variables are presented as relative frequencies (percents). Univariate analyses of categorical variables were performed by using Fisher exact test (two-tailed P value) for group comparisons between patients with type II endoleaks and evidence of
JOURNAL OF VASCULAR SURGERY Volume 39, Number 6
aneurysm growth and those with stable or shrinking aneurysms. Morphologic changes in aneurysm dimension were classified according to the reporting standards for EVAR.16 Continuous variables were expressed as medians and ranges; these were analyzed with the Mann-Whitney U test for unpaired comparisons. Kaplan-Meier survival estimates were used to determine freedom from aneurysm rupture and open conversion and patient survival. Cox regression analyses and receiver operating characteristics (ROC) curves were used to assess the influence of various risk factors on aneurysm expansion. The relative risk (RR) and 95% confidence intervals (95% CIs) for different variables were estimated by Cox regression. Only univariate regression analyses were performed because of the small number of patients in the different subgroups. The risk factors assessed included gender, smoking history, hypertension, need for chronic anticoagulation, initial aneurysm size, amount of wall calcification and sac thrombus, number and type of branch vessels, size of the endoleak nidus, and endograft type. Findings were considered statistically significant if the resulting P value was less than .05. For statistical analyses, SPSS for Windows version 11.0 (SPSS Inc, Chicago, Ill) and MedCalc statistical software version 7.2.0.2 (MedCalc Software, Mariakerke, Belgium) were used. RESULTS Study patients. Devices used for EVAR in patients with persistent type II endoleaks included Zenith (n ⫽ 6; Cook, Bloomington, Ind), Excluder (n ⫽ 5; W. L. Gore & Associates, Flagstaff, Ariz), AneuRx (n ⫽ 5; Medtronic AVE, Santa Rosa, Calif), MEGS (n ⫽ 5; Montefiore Endovascular Graft System, New York, NY), Talent (n ⫽ 4; Medtronic AVE), Ancure (n ⫽ 3; Guidant, Menlo Park, Calif), Vanguard (n ⫽ 3; Boston Scientific Corp, Natick, Mass), and Quantum (n ⫽ 1; Cordis Endovascular, Miami, Fla). Thirteen patients (41%) with persistent type II endoleaks after EVAR exhibited enlargement in aneurysm diameter (by 5 mm or more), whereas 19 (59%) exhibited no significant change in aneurysm dimension or shrinkage. The median increase in aneurysm diameter in patients with significant aneurysm enlargement was 10 mm (range, 5-32 mm), which occurred at a mean of 9.7 ⫾ 6.4 months after EVAR. Of the 19 patients without aneurysm enlargement, 6 (19 %) exhibited significant reduction in sac diameter by 5 mm or more (median reduction, 13 mm; range, 5-20 mm). There were no significant differences in clinical characteristics and comorbidities between patients with persistent type II endoleaks and aneurysm enlargement and those with stable or shrinking aneurysms (Table I). The median follow-up period was similar, 31 months (range, 12-62 months) for enlarging aneurysms and 25 months (range, 6-88 months) for stable or shrinking aneurysms (P ⫽ .12). Initial aneurysm morphology and EVAR procedures. Proximal neck diameter and length were not significantly different between the two groups (Table II). The median maximum aneurysm diameter was the same in both
Timaran et al 1159
Table I. Clinical characteristics of patients with persitent type II endoleaks: enlarging aneurysm group, and stable/ shinking aneurysm group
Median age (y) Male sex Comorbidities Hypertension Tobacco abuse Anticoagulation Coronary artery disease Diabetes mellitus Hyperlipidemia
Enlarging aneurysm (%) (n ⫽ 13)
Stable/ shrinking aneurysm (%) (n ⫽ 19)
P
79 11 (85)
77 14 (71)
.51 .39
7 (54) 2 (11) 4 (31) 10 (77) 1 (8) 5 (39)
12 (63) 8 (42) 5 (26) 10 (53) 5 (26) 10 (43)
.44 .11 .54 .15 .19 .34
groups (6 cm; range, 5-9 cm in patients with aneurysm enlargement and 4.8-10 cm in patients with stable aneurysms). The amount of thrombus and wall calcification was also comparable (Table II). The number of aortic branch vessels and patency of the inferior mesenteric artery before EVAR were not significantly different between the groups. There was no significant difference with regards to the type of endograft used for EVAR between the two groups. However, stratified analysis revealed that aneurysm enlargement was more frequent in patients with persistent type II endoleaks who had undergone procedures with Talent and Vanguard devices (RR, 3.1; 95% CI, 1.5-6.1; P ⫽ .01). Postoperative morphologic changes and endoleak characteristics. Patency of the inferior mesenteric artery as well as origin and outflow sources of the endoleak were not significantly different between the two groups (Table II). The median number of vessels feeding the endoleak was 3 (range, 2-5) in the enlarging aneurysm group and 2 (range, 1-4) in the stable/shrinking aneurysm group (P ⫽ .02). Despite this difference, Cox regression analysis failed to identify the number of vessels feeding the endoleak as a significant predictor of aneurysm enlargement (RR, 1.7; 95% CI, 0.9-3.2; P ⫽ .07). The maximum diameter of the nidus of the endoleak was significantly larger in patients with aneurysm expansion than in those with stable or shrinking aneurysms (P ⬍ .001). The median maximum nidus diameter in the enlarging aneurysm group was 23 mm (range, 13-40 mm), whereas in patients with stable/shrinking aneurysms it was 8 mm (range, 5-25 mm). Cox regression analysis revealed that the maximum diameter of the nidus was a significant predictor for aneurysm enlargement (RR, 1.12; 95% CI, 1.04-1.19; P ⫽ .001). ROC curve analysis was used to determine the optimal cutoff point for the maximum nidus diameter to predict aneurysm enlargement (Fig 2). At a cutoff value of 15 mm, the sensitivity of the maximum nidus diameter to predict aneurysm enlargement was 92.3% (95% CI, 63.9%-98.7%) with a specificity of 84.2% (95% CI, 60.4%-96.4%). A maximum nidus diameter greater than 15
JOURNAL OF VASCULAR SURGERY June 2004
1160 Timaran et al
Table II. Distribution, characteristics, and procedural factors of 32 EVAR procedures comparing enalrging aneurysm group and stable/shrinking aneurysm group Enlarging aneurysm (n ⫽ 13)
Stable/shrinking aneurysm (n ⫽ 19)
P
6 (5–9) 22 (18–26) 24 (12–40) 53.8% 69.2% 28.2%
6 (4.8–10.0) 23 (19–29) 20 (8–40) 57.9% 84.2% 24.1%
.83 .34 .29 .59 .28 .42
84.6% 15.4%
78.9% 21.1%
.99 .99
Aneurysm diameter (cm) (median [range]) Proximal neck diameter (mm) (median [range]) Proximal neck length (mm) (median [range]) Patent inferior mesenteric artery Patent hypogastric arteries Aneurysm thrombus (median) Endograft configuration Bifurcated Aortouni-iliac Endoleak channel Nidus maximum diameter (mm) (median [range]) Aortic branch vessels (number [range])
23 (13–40) 3 (2–5)
8 (5–25) 2 (1–4)
⬍.001* .02*
*Mann-Whitney U test.
mm was also a significant predictor of aneurysm enlargement by Cox regression analysis (RR, 11.1; 95% CI, 1.485.8; P ⫽ .02). The median pressure in the nidus was the same in both groups (100 mm Hg; range, 60-130 mm Hg in patients with aneurysm enlargement and 50-130 mm Hg in patients with stable aneurysms; P ⫽ not significant). However, it is important to clarify that only 5 of 19 patients with stable/ shrinking aneurysms underwent angiography and sac pressure measurements, whereas 12 of 13 patients with aneurysm enlargement had these additional investigations. Clinical outcome. For all patients with persistent type II endoleaks, freedom from aneurysm rupture rates at 1, 3, and 5 years was 100%, 92%, and 92%, whereas freedom from conversion to open aneurysm repair rates was 100%, 84%, and 75%, respectively (Kaplan-Meier). Twelve patients with persistent type II endoleaks and aneurysm enlargement underwent transarterial and/or translumbar embolization on average 12 months after EVAR. Four patients without significant aneurysm enlargement also underwent coil embolization. The results of these procedures are described in detail elsewhere.2 Briefly, technical success for translumbar embolization was 71% (10 of 14), whereas for transfemoral embolization it was 38% (3 of 8). All patients with successful transarterial or translumbar embolization, ie, with resolution of the endoleak, had stabilization or shrinkage of the aneurysm sac. Three patients had persistent type II endoleaks after failed translumbar and/or translumbar embolization. One of these patients collapsed and died suddenly without a precise diagnosis. A presumed aneurysm rupture was not confirmed because a postmortem examination was not performed. The other two patients with persistent type II endoleaks after failed translumbar embolization had no further aneurysm enlargement. One aneurysm rupture, retroperitoneal and intraperitoneal, occurred in a patient with no previous treatment for his persistent type II endoleak. After EVAR, his aneurysm diameter was 6 cm, whereas the maximum diameter of the endoleak nidus was
16 mm. The aneurysm remained stable for approximately 24 months, when significant enlargement occurred (26 mm during a period of 3 months). An open procedure without explantation of the endograft for aneurysm rupture was necessary. When the aneurysm was opened, backbleeding from four lumbar arteries was observed, confirming the diagnosis of a type II endoleak. The branches were oversewn from within the sac. The patient had an uneventful recovery. Overall, long-term survival was 95% at 1 year, 61% at 3 years, and 41% at 5 years. Nine patients died during followup. No deaths were caused by confirmed aneurysm rupture in this series, although there was one death from a suspected aneurysm rupture, which could not be confirmed because a postmortem examination was not performed. DISCUSSION The results of our study indicate that the maximum diameter of the endoleak cavity or nidus was a significant predictor for aneurysm enlargement in patients with persistent type II endoleaks after EVAR. Because it is currently recognized that type II endoleaks are not always benign, particularly when associated with aneurysm expansion, our findings provide an important guide for monitoring and determining the need for additional interventions in patients with persistent type II endoleaks. Moreover, our data also support a selective approach toward patients with type II endoleaks that persist after 6 months in the setting of an aneurysm sac that is neither enlarging nor shrinking. These stable aneurysms might be safely observed when the maximum diameter of the endoleak nidus is smaller than 15 mm because the likelihood of subsequent aneurysm enlargement is low, even though there is no spontaneous thrombosis of the endoleak cavity. Conversely, a maximum endoleak nidus diameter greater than 15 mm is associated with a greater than 10-fold increased risk of aneurysm expansion, thereby justifying a more aggressive surveillance with imaging studies obtained at shorter intervals and earlier interventions.
JOURNAL OF VASCULAR SURGERY Volume 39, Number 6
Several treatment options are available to treat patients with persistent type II endoleaks. Both transarterial and translumbar embolizations with coils and other thrombogenic materials are the most widely used methods of intervention.2,9,18,19 The size of the endoleak nidus has been recognized as an important factor to determine the approach for embolization.2 In a previous study we reported our experience with the treatment of type II endoleaks.2,* Translumbar embolization was safe and effective in the majority of patients and therefore was considered our treatment of choice, particularly for endoleaks with a large nidus and multiple feeding vessels. The size of the endoleak nidus is, therefore, not only an important predictor for technical success with the translumbar approach but also a significant risk factor for aneurysm expansion and the need for translumbar embolization. Coil embolization with translumbar approach failed in two patients whose endoleak channels remained patent. Interestingly, the aneurysm sac of both patients has remained stable since translumbar embolization was performed. The size of the endoleak channel decreased in size significantly after coil embolization, with obliteration of the nidus as demonstrated by follow-up CT scans. This observation further supports the premise of a direct relationship between the size of the endoleak channel and the likelihood of aneurysm expansion. Moreover, if the same observation is consistently observed in other patients, it would suggest that the need for complete obliteration of the endoleak cavity by using translumbar embolization, although desirable, might not be necessary to prevent further aneurysm enlargement. Although the clinical significance of the maximum diameter of the endoleak cavity or nidus as a significant predictor for aneurysm enlargement after EVAR was supported by the analysis of our data with ROC curve and Cox regression models, several limitations and concerns should be considered with regard to our results and their potential applicability. First, all 6 patients with aneurysm shrinkage and 8 patients with stable aneurysms did not undergo diagnostic angiography as part of the evaluation of their endoleaks, which is considered by some authors essential for the proper classification of all endoleaks.2,9 These patients usually had evidence of aneurysm shrinkage and therefore were not candidates for any intervention. However, axial thin-slice CT images and 3D CT reconstructions were sufficient to characterize their endoleaks further. These studies helped to identify the etiology of the endoleak, excluding attachment site leaks and leaks derived from endograft defects, ie, endoleaks type I and III, respectively. Because it is currently accepted that type II endoleaks associated with a shrinking aneurysm do not require treatment, the need for contrast angiography in these patients is no longer justified. Another limitation that might be attributed to the lack of angiographic studies in all *Ohki T, Gargiulo NJ III, Kurvers H, Cynamon J, Lipsitz EC, Suggs WD, et al. The value and limitations of translumbar and transfemoral access in the management of type II endoleaks after endovascular abdominal aortic aneurysm repair. Submitted for publication.
Timaran et al 1161
Fig 2. ROC curve analysis is usually obtained to assess the accuracy of a continuous test variable to detect a specific outcome through the graphic representation of the tradeoffs between sensitivity and specificity for every possible cutoff value. In the current study, ROC curve analysis revealed that at a cutoff value of 15 mm, the sensitivity of maximum nidus diameter to detect aneurysm enlargement was 92.3% (95% CI, 63.9%-98.7%) with a specificity of 84.2% (95% CI, 60.4%-96.4%). Moreover, the proximity of the ROC curve (bold line) to the left upper corner of the graph indicates that the maximum diameter of the nidus is an accurate predictor of aneurysm enlargement; the opposite occurs when the curve comes closer to the 45-degree diagonal of the ROC space (dotted line), which indicates that a test or predictor is less accurate.
patients is that the number of vessels feeding the endoleak could be overestimated in the enlarging aneurysm group because of the greater sensitivity of angiography and CT combined, because most patients in the stable/shrinking group only underwent CT scans. The maximum diameter of the main cavity of the endoleak or nidus was measured on images obtained with thin-cut (1 to 3 mm) spiral CT scanning with delayed imaging by using calipers or CT imaging analysis software with magnification and precise cursor placement. Although the endoleak channel was often complex and irregular, the maximum diameter of the main endoleak cavity or nidus could be determined in all patients, particularly those associated with aneurysm enlargement. Aneurysm sac volume and volume of the endoleak channel were determined in only seven patients in our study. Because changes in aneurysm size occur in three dimensions, ideally both sac maximum diameter and volume should be considered to define changes in aneurysm size.16 It is likely that patients with significant aneurysm sac volume expansion could have relatively minor diameter shifts of less than 5 mm and thus were included in the stable/shrinking aneurysm group. However, aneurysm sac diameter is still the most widely used parameter to measure changes in aneurysm size and to
1162 Timaran et al
assess the risk of rupture.16,20-22 Other methods of monitoring changes in aneurysm size, including total aneurysm sac volume, have not been validated, and their use might not be cost-effective. Moreover, the role of endoleak volume has not been previously assessed; therefore its clinical significance is unknown. The current reporting standards for EVAR do not consider endoleak volume as an important characteristic of the magnitude of endoleaks.16 Therefore, its clinical utility remains to be elucidated. Our opinion is that endoleak volume would be essential to predict the risk of aneurysm enlargement in patients with persistent endoleaks after translumbar embolization. In this particular situation it is difficult to determine the maximum diameter of the nidus because this is partially obliterated with coils and other thrombogenic materials. Finally, our study included a small number of patients, which prevented us from obtaining further stratified and multivariate analyses that are necessary to confirm the independence of any predictor of aneurysm enlargement. Moreover, only a limited number of variables were analyzed to prevent overinterpretation of our data. Univariate analyses with different methods, however, confirmed the statistically significant association between the maximum diameter of the nidus and aneurysm enlargement. Despite these methodologic limitations, our study represents one of the largest series of patients with persistent type II endoleaks from a single institution. Of importance, our experience included long-term follow-up, which might account for the instances of aneurysm enlargement and rupture seen in our series. In conclusion, our data suggest that the maximum diameter of the endoleak channel is an important predictor of aneurysm enlargement in patients with persistent type II endoleaks. A diameter of the endoleak nidus greater than 15 mm is particularly significant because it is associated with a greater than 10-fold increased risk of aneurysm expansion, thereby indicating the need for more aggressive surveillance and possibly earlier intervention. REFERENCES 1. Veith FJ, Baum RA, Ohki T, Amor M, Adiseshiah M, Blankensteijn JD, et al. Nature and significance of endoleaks and endotension: summary of opinions expressed at an international conference. J Vasc Surg 2002;35: 1029-35. 2. Rhee SJ, Ohki T, Veith FJ, Kurvers H. Current status of management of type II endoleaks after endovascular repair of abdominal aortic aneurysms. Ann Vasc Surg 2003;17:335-44. 3. Gorich J, Rilinger N, Sokiranski R, Kramer S, Schutz A, SunderPlassmann L, et al. Embolization of type II endoleaks fed by the inferior mesenteric artery: using the superior mesenteric artery approach. J Endovasc Ther 2000;7:297-301. 4. Politz JK, Newman VS, Stewart MT. Late abdominal aortic aneurysm rupture after AneuRx repair: a report of three cases. J Vasc Surg 2000;31:599-606.
JOURNAL OF VASCULAR SURGERY June 2004
5. 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;5:189-91. 6. Harris PL, Vallabhaneni SR, Desgranges P, Becquemin JP, van Marrewijk C, Laheij RJ. Incidence and risk factors of late rupture, conversion, and death after endovascular repair of infrarenal aortic aneurysms: the EUROSTAR experience—European Collaborators on Stent/graft techniques for aortic aneurysm repair. J Vasc Surg 2000;32:739-49. 7. Chuter TA, Faruqi RM, Sawhney R, Reilly LM, Kerlan RB, Canto CJ, et al. Endoleak after endovascular repair of abdominal aortic aneurysm. J Vasc Surg 2001;34:98-105. 8. van Marrewijk C, Buth J, Harris PL, Norgren L, Nevelsteen A, Wyatt MG. Significance of endoleaks after endovascular repair of abdominal aortic aneurysms: The EUROSTAR experience. J Vasc Surg 2002;35: 461-73. 9. Baum RA, Carpenter JP, Golden MA, Velazquez OC, Clark TW, Stavropoulos SW, et al. Treatment of type 2 endoleaks after endovascular repair of abdominal aortic aneurysms: comparison of transarterial and translumbar techniques. J Vasc Surg 2002;35:23-9. 10. Zarins CK, White RA, Hodgson KJ, Schwarten D, Fogarty TJ. Endoleak as a predictor of outcome after endovascular aneurysm repair: AneuRx multicenter clinical trial. J Vasc Surg 2000;32:90-107. 11. Baum RA, Carpenter JP, Cope C, Golden MA, Velazquez OC, Neschis DG, et al. Aneurysm sac pressure measurements after endovascular repair of abdominal aortic aneurysms. J Vasc Surg 2001;33:32-41. 12. Hinchliffe RJ, Singh-Ranger R, Davidson IR, Hopkinson BR. Rupture of an abdominal aortic aneurysm secondary to type II endoleak. Eur J Vasc Endovasc Surg 2001;22:563-5. 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;33:1271-4. 14. Bernhard VM, Mitchell RS, Matsumura JS, Brewster DC, Decker M, Lamparello P, et al. Ruptured abdominal aortic aneurysm after endovascular repair. J Vasc Surg 2002;35:1155-62. 15. Chaikof EL, Fillinger MF, Matsumura JS, Rutherford RB, White GH, Blankensteijn JD, et al. Identifying and grading factors that modify the outcome of endovascular aortic aneurysm repair. J Vasc Surg 2002;35: 1061-6. 16. Chaikof EL, Blankensteijn JD, Harris PL, White GH, Zarins CK, Bernhard VM, et al. Reporting standards for endovascular aortic aneurysm repair. J Vasc Surg 2002;35:1048-60. 17. Cayne NC, Veith FJ, Lipsitz EC, Ohki T, Mehta M, Gargiulo NJ III, et al. Variability of maximal aortic aneurysm diameter measurements on CT scan: significance and methods to minimize. J Vasc Surg. In press 2004. 18. Solis MM, Ayerdi J, Babcock GA, Parra JR, McLafferty RB, Gruneiro LA, et al. Mechanism of failure in the treatment of type II endoleak with percutaneous coil embolization. J Vasc Surg 2002;36:485-91. 19. Kasirajan K, Matteson B, Marek JM, Langsfeld M. Technique and results of transfemoral superselective coil embolization of type II lumbar endoleak. J Vasc Surg 2003;38:61-6. 20. Cronenwett JL, Murphy TF, Zelenock GB, Whitehouse WM Jr, Lindenauer SM, Graham LM, et al. Actuarial analysis of variables associated with rupture of small abdominal aortic aneurysms. Surgery 1985;98: 472-83. 21. Brown PM, Pattenden R, Vernooy C, Zelt DT, Gutelius JR. Selective management of abdominal aortic aneurysms in a prospective measurement program. J Vasc Surg 1996;23:213-20. 22. Brown PM, Zelt DT, Sobolev B. The risk of rupture in untreated aneurysms: the impact of size, gender, and expansion rate. J Vasc Surg 2003;37:280-4. Submitted Sep 18, 2003; accepted Dec 28, 2003. Available online Apr 2, 2003.
Predicting aneurysm enlargement in patients with persistent type II endoleaks Carlos H. Timaran, MD, Takao Ohki, MD, PhD, Soo J. Rhee, MD, Frank J. Veith, MD, Nicholas J. Gargiulo III, MD, Hisako Toriumi, MD, Mahmood B. Malas, MD, William D. Suggs, MD, Reese A. Wain, MD, and Evan C. Lipsitz, MD, New York, NY Objective: The clinical significance of type II endoleaks is not well understood. Some evidence, however, indicates that some type II endoleaks might result in aneurysm enlargement and rupture. To identify factors that might contribute to aneurysm expansion, we analyzed the influence of several variables on aneurysm growth in patients with persistent type II endoleaks after endovascular aortic aneurysm repair (EVAR). Methods: In a series of 348 EVARs performed during a 10-year period, 32 patients (9.2%) developed type II endoleaks that persisted for more than 6 months. Variables analyzed included those defined by the reporting standards for EVAR (SVS/AAVS) as well as other endoleak characteristics. Univariate, receiver operating characteristic curve, and Cox regression analyses were used to determine the association between variables and aneurysm enlargement. Results: The median follow-up period was 26.5 months (range, 6-88 months). Thirteen patients (41%) had aneurysm enlargement by 5 mm or more (median increase in diameter, 10 mm), whereas 19 (59%) had stable or shrinking aneurysm diameter. Univariate and Cox regression analyses identified the maximum diameter of the endoleak cavity, ie, the size of the nidus as defined on contrast computed tomography scan, as a significant predictor for aneurysm enlargement (relative risk, 1.12; 95% confidence interval, 1.04-1.19; P ⴝ .001). The median size of the nidus was 23 mm (range, 13-40 mm) in patients with aneurysm enlargement and 8 mm (range, 5-25 mm) in those without expansion (Mann-Whitney U test, P < .001). Moreover, receiver operating characteristic curve and Cox regression analyses showed that a maximum nidus diameter greater than 15 mm was particularly associated with an increased risk of aneurysm enlargement (relative risk, 11.1; 95% confidence interval, 1.4-85.8; P ⴝ .02). Other risk factors including gender, smoking history, hypertension, need of anticoagulation, aneurysm diameter, type of endograft used, and number or type of collateral vessels were not significant predictors of aneurysm enlargement. Conclusions: In patients with persistent type II endoleaks after EVAR, the maximum diameter of the endoleak cavity or nidus is an important predictor of aneurysm growth and might indicate the need for more aggressive surveillance as well as earlier treatment. (J Vasc Surg 2004;39:1157-62.)
The most frequent mechanism of failure after endovascular aortic aneurysm repair (EVAR) is the occurrence of an endoleak, ie, the persistence of blood flow in the aneurysm sac outside the endograft.1,2 Type II endoleaks are attributed to retrograde flow from patent lumbar arteries, inferior mesenteric artery, and other collateral aortic branches.2-4 Although type II endoleaks are frequently evident in intraoperative angiograms after EVAR, the majority of these endoleaks resolve as a result of the reversal of anticoagulation and the subsequent thrombosis of the aneurysm sac and its side branches.2 Persistent type II endoleaks develop in approximately 5% to 25% of the patients after EVAR.1,2,5-9 The exact clinical significance of persistent type II endoleaks is not defined because these might be associated with aneurysm enlargement, stability, and From the Division of Vascular Surgery, Department of Surgery, Montefiore Medical Center, Albert Einstein College of Medicine. Competition of interest: none. Presented at the Thirtieth Anniversary Meeting of the New England Society for Vascular Surgery, Newport, RI, Sep 19-21, 2003. Winning paper of the Darling Award of the New England Society for Vascular Surgery. Reprint requests: Takao Ohki, MD, Montefiore Medical Center, 111E 210th St, Bronx, NY 10467 (e-mail: [email protected]). 0741-5214/$30.00 Copyright © 2004 by The Society for Vascular Surgery. doi:10.1016/j.jvs.2003.12.033
shrinkage.1,2,8,10 Therefore, the optimal method of diagnosis and treatment of type II endoleaks is also controversial.9 As more experience with the diagnosis and treatment of type II endoleaks has been reported, it is now well recognized that type II endoleaks are not always benign. For instance, recent studies have documented that type II endoleaks might produce systemic blood pressures within the aneurysm sac that might result in aneurysm enlargement and rupture.4,8,11-14 Because the risk of aneurysm rupture is related to the maximum diameter of the aneurysm, general agreement exists that persistent type II endoleaks associated with aneurysm enlargement require aggressive management. However, it is still unknown which type II endoleaks will result in aneurysm expansion, thereby increasing the risk of rupture. This study was designed to assess the influence of several variables on aneurysm expansion in patients with persistent type II endoleaks after EVAR to define when a more aggressive approach of this type of endoleak is indicated. The most current recommended reporting standards were used to define the different variables.15,16 METHODS Among 348 patients who underwent EVAR during a 10-year period (November 1992 to September 2002), 32 1157
1158 Timaran et al
JOURNAL OF VASCULAR SURGERY June 2004
Fig 1. Maximum diameter measurements of the main endoleak cavity or nidus (e) was obtained from CT scan images by using calipers or CT imaging analysis software with magnification and precise cursor placement. A, Small endoleak nidus (2) measuring 9 mm. Although the endoleak channel was sometimes irregular and complex, a maximum nidus diameter could be determined in all patients. B, Large endoleak nidus measuring 23 mm in maximum diameter. When more than one nidus was seen, the maximum diameter of the largest nidus was selected.
patients (9.2%) developed type II endoleaks that were initially observed at the first postoperative month or at any time thereafter and persisted for more than 6 months. These 32 patients with persistent type II endoleaks were included in a cohort study. Patients with type II endoleaks associated with either type I or type III endoleaks at any time were specifically excluded. Patient clinical and procedural characteristics and follow-up results of patients undergoing EVAR were prospectively collected. The medical records, computed tomography (CT) scans, and angiographic studies of these patients were, however, further reviewed to define clinical characteristics, risk factors, morphologic features of endoleaks, and changes in aortic aneurysms according to the current reporting standards for EVAR prepared and revised by the Ad Hoc Committee for Standardized Reporting Practices in Vascular Surgery (SVS/AAVS).15,16 Patients with persistent type II endoleaks underwent CT scans every 3 to 6 months, and no patients were lost to follow-up. Axial thin-slice (1.5-3 mm thick) CT imaging and occasionally three-dimensional (3D) CT reconstruction models were obtained to confirm the presence of an endoleak and to further define the etiology and the most likely classification of the leak. Sequential changes in aneurysm size were determined with serial measurements of the maximum aneurysm diameter by using a standardized method previously described to decrease interobserver variability.17 Patients with persistent type II endoleaks who had documented aneurysm enlargement on follow-up CT scanning underwent diagnostic
angiography to characterize the endoleak. Once a type II endoleak was confirmed angiographically in patients with evidence of aneurysm expansion, additional treatment with either transarterial or translumbar embolization was performed. The technique of these treatment modalities has been described in detail before. Follow-up dynamic CT scans with 3-mm slice-thickness were obtained at 3, 6, and 12 months and annually thereafter. Embolization was considered to be technically successful when resolution of the endoleak was obtained with stabilization or shrinkage of the aneurysm sac. All follow-up CT scans and angiograms of patients with persistent type II endoleaks were reviewed to characterize the endoleak channel in greater detail. Endoleak angiography and pressure measurements were available in patients who underwent translumbar embolization. The number, origin, and outflow vessels feeding the endoleak were specified as suggested by the current reporting standards.15 The size of the main endoleak cavity or nidus was recorded including the maximum diameter, short axis, and length of the endoleak cavity. Calipers or CT imaging analysis software with magnification and precise cursor placement were used for diameter measurements obtained from CT scan images (Fig 1). Descriptive statistics for categorical variables are presented as relative frequencies (percents). Univariate analyses of categorical variables were performed by using Fisher exact test (two-tailed P value) for group comparisons between patients with type II endoleaks and evidence of
JOURNAL OF VASCULAR SURGERY Volume 39, Number 6
aneurysm growth and those with stable or shrinking aneurysms. Morphologic changes in aneurysm dimension were classified according to the reporting standards for EVAR.16 Continuous variables were expressed as medians and ranges; these were analyzed with the Mann-Whitney U test for unpaired comparisons. Kaplan-Meier survival estimates were used to determine freedom from aneurysm rupture and open conversion and patient survival. Cox regression analyses and receiver operating characteristics (ROC) curves were used to assess the influence of various risk factors on aneurysm expansion. The relative risk (RR) and 95% confidence intervals (95% CIs) for different variables were estimated by Cox regression. Only univariate regression analyses were performed because of the small number of patients in the different subgroups. The risk factors assessed included gender, smoking history, hypertension, need for chronic anticoagulation, initial aneurysm size, amount of wall calcification and sac thrombus, number and type of branch vessels, size of the endoleak nidus, and endograft type. Findings were considered statistically significant if the resulting P value was less than .05. For statistical analyses, SPSS for Windows version 11.0 (SPSS Inc, Chicago, Ill) and MedCalc statistical software version 7.2.0.2 (MedCalc Software, Mariakerke, Belgium) were used. RESULTS Study patients. Devices used for EVAR in patients with persistent type II endoleaks included Zenith (n ⫽ 6; Cook, Bloomington, Ind), Excluder (n ⫽ 5; W. L. Gore & Associates, Flagstaff, Ariz), AneuRx (n ⫽ 5; Medtronic AVE, Santa Rosa, Calif), MEGS (n ⫽ 5; Montefiore Endovascular Graft System, New York, NY), Talent (n ⫽ 4; Medtronic AVE), Ancure (n ⫽ 3; Guidant, Menlo Park, Calif), Vanguard (n ⫽ 3; Boston Scientific Corp, Natick, Mass), and Quantum (n ⫽ 1; Cordis Endovascular, Miami, Fla). Thirteen patients (41%) with persistent type II endoleaks after EVAR exhibited enlargement in aneurysm diameter (by 5 mm or more), whereas 19 (59%) exhibited no significant change in aneurysm dimension or shrinkage. The median increase in aneurysm diameter in patients with significant aneurysm enlargement was 10 mm (range, 5-32 mm), which occurred at a mean of 9.7 ⫾ 6.4 months after EVAR. Of the 19 patients without aneurysm enlargement, 6 (19 %) exhibited significant reduction in sac diameter by 5 mm or more (median reduction, 13 mm; range, 5-20 mm). There were no significant differences in clinical characteristics and comorbidities between patients with persistent type II endoleaks and aneurysm enlargement and those with stable or shrinking aneurysms (Table I). The median follow-up period was similar, 31 months (range, 12-62 months) for enlarging aneurysms and 25 months (range, 6-88 months) for stable or shrinking aneurysms (P ⫽ .12). Initial aneurysm morphology and EVAR procedures. Proximal neck diameter and length were not significantly different between the two groups (Table II). The median maximum aneurysm diameter was the same in both
Timaran et al 1159
Table I. Clinical characteristics of patients with persitent type II endoleaks: enlarging aneurysm group, and stable/ shinking aneurysm group
Median age (y) Male sex Comorbidities Hypertension Tobacco abuse Anticoagulation Coronary artery disease Diabetes mellitus Hyperlipidemia
Enlarging aneurysm (%) (n ⫽ 13)
Stable/ shrinking aneurysm (%) (n ⫽ 19)
P
79 11 (85)
77 14 (71)
.51 .39
7 (54) 2 (11) 4 (31) 10 (77) 1 (8) 5 (39)
12 (63) 8 (42) 5 (26) 10 (53) 5 (26) 10 (43)
.44 .11 .54 .15 .19 .34
groups (6 cm; range, 5-9 cm in patients with aneurysm enlargement and 4.8-10 cm in patients with stable aneurysms). The amount of thrombus and wall calcification was also comparable (Table II). The number of aortic branch vessels and patency of the inferior mesenteric artery before EVAR were not significantly different between the groups. There was no significant difference with regards to the type of endograft used for EVAR between the two groups. However, stratified analysis revealed that aneurysm enlargement was more frequent in patients with persistent type II endoleaks who had undergone procedures with Talent and Vanguard devices (RR, 3.1; 95% CI, 1.5-6.1; P ⫽ .01). Postoperative morphologic changes and endoleak characteristics. Patency of the inferior mesenteric artery as well as origin and outflow sources of the endoleak were not significantly different between the two groups (Table II). The median number of vessels feeding the endoleak was 3 (range, 2-5) in the enlarging aneurysm group and 2 (range, 1-4) in the stable/shrinking aneurysm group (P ⫽ .02). Despite this difference, Cox regression analysis failed to identify the number of vessels feeding the endoleak as a significant predictor of aneurysm enlargement (RR, 1.7; 95% CI, 0.9-3.2; P ⫽ .07). The maximum diameter of the nidus of the endoleak was significantly larger in patients with aneurysm expansion than in those with stable or shrinking aneurysms (P ⬍ .001). The median maximum nidus diameter in the enlarging aneurysm group was 23 mm (range, 13-40 mm), whereas in patients with stable/shrinking aneurysms it was 8 mm (range, 5-25 mm). Cox regression analysis revealed that the maximum diameter of the nidus was a significant predictor for aneurysm enlargement (RR, 1.12; 95% CI, 1.04-1.19; P ⫽ .001). ROC curve analysis was used to determine the optimal cutoff point for the maximum nidus diameter to predict aneurysm enlargement (Fig 2). At a cutoff value of 15 mm, the sensitivity of the maximum nidus diameter to predict aneurysm enlargement was 92.3% (95% CI, 63.9%-98.7%) with a specificity of 84.2% (95% CI, 60.4%-96.4%). A maximum nidus diameter greater than 15
JOURNAL OF VASCULAR SURGERY June 2004
1160 Timaran et al
Table II. Distribution, characteristics, and procedural factors of 32 EVAR procedures comparing enalrging aneurysm group and stable/shrinking aneurysm group Enlarging aneurysm (n ⫽ 13)
Stable/shrinking aneurysm (n ⫽ 19)
P
6 (5–9) 22 (18–26) 24 (12–40) 53.8% 69.2% 28.2%
6 (4.8–10.0) 23 (19–29) 20 (8–40) 57.9% 84.2% 24.1%
.83 .34 .29 .59 .28 .42
84.6% 15.4%
78.9% 21.1%
.99 .99
Aneurysm diameter (cm) (median [range]) Proximal neck diameter (mm) (median [range]) Proximal neck length (mm) (median [range]) Patent inferior mesenteric artery Patent hypogastric arteries Aneurysm thrombus (median) Endograft configuration Bifurcated Aortouni-iliac Endoleak channel Nidus maximum diameter (mm) (median [range]) Aortic branch vessels (number [range])
23 (13–40) 3 (2–5)
8 (5–25) 2 (1–4)
⬍.001* .02*
*Mann-Whitney U test.
mm was also a significant predictor of aneurysm enlargement by Cox regression analysis (RR, 11.1; 95% CI, 1.485.8; P ⫽ .02). The median pressure in the nidus was the same in both groups (100 mm Hg; range, 60-130 mm Hg in patients with aneurysm enlargement and 50-130 mm Hg in patients with stable aneurysms; P ⫽ not significant). However, it is important to clarify that only 5 of 19 patients with stable/ shrinking aneurysms underwent angiography and sac pressure measurements, whereas 12 of 13 patients with aneurysm enlargement had these additional investigations. Clinical outcome. For all patients with persistent type II endoleaks, freedom from aneurysm rupture rates at 1, 3, and 5 years was 100%, 92%, and 92%, whereas freedom from conversion to open aneurysm repair rates was 100%, 84%, and 75%, respectively (Kaplan-Meier). Twelve patients with persistent type II endoleaks and aneurysm enlargement underwent transarterial and/or translumbar embolization on average 12 months after EVAR. Four patients without significant aneurysm enlargement also underwent coil embolization. The results of these procedures are described in detail elsewhere.2 Briefly, technical success for translumbar embolization was 71% (10 of 14), whereas for transfemoral embolization it was 38% (3 of 8). All patients with successful transarterial or translumbar embolization, ie, with resolution of the endoleak, had stabilization or shrinkage of the aneurysm sac. Three patients had persistent type II endoleaks after failed translumbar and/or translumbar embolization. One of these patients collapsed and died suddenly without a precise diagnosis. A presumed aneurysm rupture was not confirmed because a postmortem examination was not performed. The other two patients with persistent type II endoleaks after failed translumbar embolization had no further aneurysm enlargement. One aneurysm rupture, retroperitoneal and intraperitoneal, occurred in a patient with no previous treatment for his persistent type II endoleak. After EVAR, his aneurysm diameter was 6 cm, whereas the maximum diameter of the endoleak nidus was
16 mm. The aneurysm remained stable for approximately 24 months, when significant enlargement occurred (26 mm during a period of 3 months). An open procedure without explantation of the endograft for aneurysm rupture was necessary. When the aneurysm was opened, backbleeding from four lumbar arteries was observed, confirming the diagnosis of a type II endoleak. The branches were oversewn from within the sac. The patient had an uneventful recovery. Overall, long-term survival was 95% at 1 year, 61% at 3 years, and 41% at 5 years. Nine patients died during followup. No deaths were caused by confirmed aneurysm rupture in this series, although there was one death from a suspected aneurysm rupture, which could not be confirmed because a postmortem examination was not performed. DISCUSSION The results of our study indicate that the maximum diameter of the endoleak cavity or nidus was a significant predictor for aneurysm enlargement in patients with persistent type II endoleaks after EVAR. Because it is currently recognized that type II endoleaks are not always benign, particularly when associated with aneurysm expansion, our findings provide an important guide for monitoring and determining the need for additional interventions in patients with persistent type II endoleaks. Moreover, our data also support a selective approach toward patients with type II endoleaks that persist after 6 months in the setting of an aneurysm sac that is neither enlarging nor shrinking. These stable aneurysms might be safely observed when the maximum diameter of the endoleak nidus is smaller than 15 mm because the likelihood of subsequent aneurysm enlargement is low, even though there is no spontaneous thrombosis of the endoleak cavity. Conversely, a maximum endoleak nidus diameter greater than 15 mm is associated with a greater than 10-fold increased risk of aneurysm expansion, thereby justifying a more aggressive surveillance with imaging studies obtained at shorter intervals and earlier interventions.
JOURNAL OF VASCULAR SURGERY Volume 39, Number 6
Several treatment options are available to treat patients with persistent type II endoleaks. Both transarterial and translumbar embolizations with coils and other thrombogenic materials are the most widely used methods of intervention.2,9,18,19 The size of the endoleak nidus has been recognized as an important factor to determine the approach for embolization.2 In a previous study we reported our experience with the treatment of type II endoleaks.2,* Translumbar embolization was safe and effective in the majority of patients and therefore was considered our treatment of choice, particularly for endoleaks with a large nidus and multiple feeding vessels. The size of the endoleak nidus is, therefore, not only an important predictor for technical success with the translumbar approach but also a significant risk factor for aneurysm expansion and the need for translumbar embolization. Coil embolization with translumbar approach failed in two patients whose endoleak channels remained patent. Interestingly, the aneurysm sac of both patients has remained stable since translumbar embolization was performed. The size of the endoleak channel decreased in size significantly after coil embolization, with obliteration of the nidus as demonstrated by follow-up CT scans. This observation further supports the premise of a direct relationship between the size of the endoleak channel and the likelihood of aneurysm expansion. Moreover, if the same observation is consistently observed in other patients, it would suggest that the need for complete obliteration of the endoleak cavity by using translumbar embolization, although desirable, might not be necessary to prevent further aneurysm enlargement. Although the clinical significance of the maximum diameter of the endoleak cavity or nidus as a significant predictor for aneurysm enlargement after EVAR was supported by the analysis of our data with ROC curve and Cox regression models, several limitations and concerns should be considered with regard to our results and their potential applicability. First, all 6 patients with aneurysm shrinkage and 8 patients with stable aneurysms did not undergo diagnostic angiography as part of the evaluation of their endoleaks, which is considered by some authors essential for the proper classification of all endoleaks.2,9 These patients usually had evidence of aneurysm shrinkage and therefore were not candidates for any intervention. However, axial thin-slice CT images and 3D CT reconstructions were sufficient to characterize their endoleaks further. These studies helped to identify the etiology of the endoleak, excluding attachment site leaks and leaks derived from endograft defects, ie, endoleaks type I and III, respectively. Because it is currently accepted that type II endoleaks associated with a shrinking aneurysm do not require treatment, the need for contrast angiography in these patients is no longer justified. Another limitation that might be attributed to the lack of angiographic studies in all *Ohki T, Gargiulo NJ III, Kurvers H, Cynamon J, Lipsitz EC, Suggs WD, et al. The value and limitations of translumbar and transfemoral access in the management of type II endoleaks after endovascular abdominal aortic aneurysm repair. Submitted for publication.
Timaran et al 1161
Fig 2. ROC curve analysis is usually obtained to assess the accuracy of a continuous test variable to detect a specific outcome through the graphic representation of the tradeoffs between sensitivity and specificity for every possible cutoff value. In the current study, ROC curve analysis revealed that at a cutoff value of 15 mm, the sensitivity of maximum nidus diameter to detect aneurysm enlargement was 92.3% (95% CI, 63.9%-98.7%) with a specificity of 84.2% (95% CI, 60.4%-96.4%). Moreover, the proximity of the ROC curve (bold line) to the left upper corner of the graph indicates that the maximum diameter of the nidus is an accurate predictor of aneurysm enlargement; the opposite occurs when the curve comes closer to the 45-degree diagonal of the ROC space (dotted line), which indicates that a test or predictor is less accurate.
patients is that the number of vessels feeding the endoleak could be overestimated in the enlarging aneurysm group because of the greater sensitivity of angiography and CT combined, because most patients in the stable/shrinking group only underwent CT scans. The maximum diameter of the main cavity of the endoleak or nidus was measured on images obtained with thin-cut (1 to 3 mm) spiral CT scanning with delayed imaging by using calipers or CT imaging analysis software with magnification and precise cursor placement. Although the endoleak channel was often complex and irregular, the maximum diameter of the main endoleak cavity or nidus could be determined in all patients, particularly those associated with aneurysm enlargement. Aneurysm sac volume and volume of the endoleak channel were determined in only seven patients in our study. Because changes in aneurysm size occur in three dimensions, ideally both sac maximum diameter and volume should be considered to define changes in aneurysm size.16 It is likely that patients with significant aneurysm sac volume expansion could have relatively minor diameter shifts of less than 5 mm and thus were included in the stable/shrinking aneurysm group. However, aneurysm sac diameter is still the most widely used parameter to measure changes in aneurysm size and to
1162 Timaran et al
assess the risk of rupture.16,20-22 Other methods of monitoring changes in aneurysm size, including total aneurysm sac volume, have not been validated, and their use might not be cost-effective. Moreover, the role of endoleak volume has not been previously assessed; therefore its clinical significance is unknown. The current reporting standards for EVAR do not consider endoleak volume as an important characteristic of the magnitude of endoleaks.16 Therefore, its clinical utility remains to be elucidated. Our opinion is that endoleak volume would be essential to predict the risk of aneurysm enlargement in patients with persistent endoleaks after translumbar embolization. In this particular situation it is difficult to determine the maximum diameter of the nidus because this is partially obliterated with coils and other thrombogenic materials. Finally, our study included a small number of patients, which prevented us from obtaining further stratified and multivariate analyses that are necessary to confirm the independence of any predictor of aneurysm enlargement. Moreover, only a limited number of variables were analyzed to prevent overinterpretation of our data. Univariate analyses with different methods, however, confirmed the statistically significant association between the maximum diameter of the nidus and aneurysm enlargement. Despite these methodologic limitations, our study represents one of the largest series of patients with persistent type II endoleaks from a single institution. Of importance, our experience included long-term follow-up, which might account for the instances of aneurysm enlargement and rupture seen in our series. In conclusion, our data suggest that the maximum diameter of the endoleak channel is an important predictor of aneurysm enlargement in patients with persistent type II endoleaks. A diameter of the endoleak nidus greater than 15 mm is particularly significant because it is associated with a greater than 10-fold increased risk of aneurysm expansion, thereby indicating the need for more aggressive surveillance and possibly earlier intervention. REFERENCES 1. Veith FJ, Baum RA, Ohki T, Amor M, Adiseshiah M, Blankensteijn JD, et al. Nature and significance of endoleaks and endotension: summary of opinions expressed at an international conference. J Vasc Surg 2002;35: 1029-35. 2. Rhee SJ, Ohki T, Veith FJ, Kurvers H. Current status of management of type II endoleaks after endovascular repair of abdominal aortic aneurysms. Ann Vasc Surg 2003;17:335-44. 3. Gorich J, Rilinger N, Sokiranski R, Kramer S, Schutz A, SunderPlassmann L, et al. Embolization of type II endoleaks fed by the inferior mesenteric artery: using the superior mesenteric artery approach. J Endovasc Ther 2000;7:297-301. 4. Politz JK, Newman VS, Stewart MT. Late abdominal aortic aneurysm rupture after AneuRx repair: a report of three cases. J Vasc Surg 2000;31:599-606.
JOURNAL OF VASCULAR SURGERY June 2004
5. 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;5:189-91. 6. Harris PL, Vallabhaneni SR, Desgranges P, Becquemin JP, van Marrewijk C, Laheij RJ. Incidence and risk factors of late rupture, conversion, and death after endovascular repair of infrarenal aortic aneurysms: the EUROSTAR experience—European Collaborators on Stent/graft techniques for aortic aneurysm repair. J Vasc Surg 2000;32:739-49. 7. Chuter TA, Faruqi RM, Sawhney R, Reilly LM, Kerlan RB, Canto CJ, et al. Endoleak after endovascular repair of abdominal aortic aneurysm. J Vasc Surg 2001;34:98-105. 8. van Marrewijk C, Buth J, Harris PL, Norgren L, Nevelsteen A, Wyatt MG. Significance of endoleaks after endovascular repair of abdominal aortic aneurysms: The EUROSTAR experience. J Vasc Surg 2002;35: 461-73. 9. Baum RA, Carpenter JP, Golden MA, Velazquez OC, Clark TW, Stavropoulos SW, et al. Treatment of type 2 endoleaks after endovascular repair of abdominal aortic aneurysms: comparison of transarterial and translumbar techniques. J Vasc Surg 2002;35:23-9. 10. Zarins CK, White RA, Hodgson KJ, Schwarten D, Fogarty TJ. Endoleak as a predictor of outcome after endovascular aneurysm repair: AneuRx multicenter clinical trial. J Vasc Surg 2000;32:90-107. 11. Baum RA, Carpenter JP, Cope C, Golden MA, Velazquez OC, Neschis DG, et al. Aneurysm sac pressure measurements after endovascular repair of abdominal aortic aneurysms. J Vasc Surg 2001;33:32-41. 12. Hinchliffe RJ, Singh-Ranger R, Davidson IR, Hopkinson BR. Rupture of an abdominal aortic aneurysm secondary to type II endoleak. Eur J Vasc Endovasc Surg 2001;22:563-5. 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;33:1271-4. 14. Bernhard VM, Mitchell RS, Matsumura JS, Brewster DC, Decker M, Lamparello P, et al. Ruptured abdominal aortic aneurysm after endovascular repair. J Vasc Surg 2002;35:1155-62. 15. Chaikof EL, Fillinger MF, Matsumura JS, Rutherford RB, White GH, Blankensteijn JD, et al. Identifying and grading factors that modify the outcome of endovascular aortic aneurysm repair. J Vasc Surg 2002;35: 1061-6. 16. Chaikof EL, Blankensteijn JD, Harris PL, White GH, Zarins CK, Bernhard VM, et al. Reporting standards for endovascular aortic aneurysm repair. J Vasc Surg 2002;35:1048-60. 17. Cayne NC, Veith FJ, Lipsitz EC, Ohki T, Mehta M, Gargiulo NJ III, et al. Variability of maximal aortic aneurysm diameter measurements on CT scan: significance and methods to minimize. J Vasc Surg. In press 2004. 18. Solis MM, Ayerdi J, Babcock GA, Parra JR, McLafferty RB, Gruneiro LA, et al. Mechanism of failure in the treatment of type II endoleak with percutaneous coil embolization. J Vasc Surg 2002;36:485-91. 19. Kasirajan K, Matteson B, Marek JM, Langsfeld M. Technique and results of transfemoral superselective coil embolization of type II lumbar endoleak. J Vasc Surg 2003;38:61-6. 20. Cronenwett JL, Murphy TF, Zelenock GB, Whitehouse WM Jr, Lindenauer SM, Graham LM, et al. Actuarial analysis of variables associated with rupture of small abdominal aortic aneurysms. Surgery 1985;98: 472-83. 21. Brown PM, Pattenden R, Vernooy C, Zelt DT, Gutelius JR. Selective management of abdominal aortic aneurysms in a prospective measurement program. J Vasc Surg 1996;23:213-20. 22. Brown PM, Zelt DT, Sobolev B. The risk of rupture in untreated aneurysms: the impact of size, gender, and expansion rate. J Vasc Surg 2003;37:280-4. Submitted Sep 18, 2003; accepted Dec 28, 2003. Available online Apr 2, 2003.