Type-II Endoleaks Following Endovascular AAA Repair: Preoperative Predictors and Long-term Effects

Transcription

1 503 VASCULAR FELLOWS FORUM 2001, FIRST PLACE Type-II Endoleaks Following Endovascular AAA Repair: Preoperative Predictors and Long-term Effects Frank R. Arko, MD; Geoffrey D. Rubin, MD; Bonnie L. Johnson, RDMS, RVT; Bradley B. Hill, MD; Thomas J. Fogarty, MD; and Christopher K. Zarins, MD Division of Vascular Surgery, Stanford University Medical Center, Stanford, California, USA Purpose: To determine the significance of persistent type-ii endoleaks and whether they can be predicted preoperatively in patients with abdominal aortic aneurysms (AAA). Methods: The charts of all AAA patients treated with the AneuRx stent-graft at a single center from 1996 to 1998 were reviewed. Patients with 12-month follow-up or type-i endoleaks were excluded. The presence or absence of type-ii endoleaks was determined from duplex imaging and computed tomographic angiography. Three groups were identified and compared: 16 patients with persistent type-ii endoleaks (PE), 14 patients with transient type-ii endoleaks (TE), and 16 patients with no endoleak (NE). Results: The groups did not differ with regard to age, preoperative comorbidities, followup time, and AAA neck diameter and length. AAA diameters were mm for NE, mm for TE, and mm for PE. The inferior mesenteric artery (IMA) was patent in 5 (31%) NE patients, 6 (43%) TE patients, and 13 (81%) PE patients (p 0.01). The number of patent lumbar arteries visualized preoperatively was in NE, in TE, and in PE (p ). Patent IMAs (RR 0.82, p0.01) and 2 lumbar arteries (RR 0.40, p ) were identified as independent preoperative risk factors for persistent endoleaks. There were no changes in mean diameter or volume in aneurysms with persistent endoleaks. Conclusions: No adverse clinical events were related to the presence of type-ii endoleaks, but there was no decrease in aneurysm size in patients with persistent type-ii leaks. Patients with a large, patent IMA, or 2 lumbar arteries on preoperative CT angiography are at higher risk for persistent type-ii endoleaks. J Endovasc Ther Key words: endovascular grafting, AneuRx stent-graft, inferior mesenteric artery, lumbar arteries, aneurysm volume Endovascular repair has significantly reduced the morbidity of abdominal aortic aneurysm (AAA) surgery; however, the long-term effec- Presented at International Congress XIV on Endovascular Interventions, Scottsdale, Arizona, USA, February 11 15, tiveness of endovascular grafting is unclear. While successful aneurysm exclusion has Address for correspondence and reprints: Christopher K. Zarins, MD, Division of Vascular Surgery, Stanford University Medical Center, 300 Pasteur Drive, Room H3600, Stanford, CA USA. Fax: ; zarins@stanford.edu 2001 by the INTERNATIONAL SOCIETY OF ENDOVASCULAR SPECIALISTS

2 504 TYPE-II ENDOLEAKS been achieved in 50% to 90% of cases, persistent flow within the aneurysm sac has been reported in 15% to 52% of patients following endovascular repair. 1 6 Although some consider that any evidence of endoleak perpetuates the risk of aneurysm rupture, the true significance of endoleak has not been determined. There is little debate that type-i endoleaks originating from attachment sites in the aneurysm neck or iliac arteries indicate incomplete repair with continued risk of aneurysm rupture These endoleaks should be repaired either with endovascular or standard open repair with little delay. However, there is no consensus as to the relevance of type-ii endoleaks, 5,6,14 which are related to retrograde flow into the aneurysm sac through the lumbar arteries or inferior mesenteric artery (IMA). The impact of type-ii endoleaks on longterm outcome following endovascular AAA repair is unknown. The purpose of this study was to determine the clinical significance of persistent type-ii endoleaks long-term and whether they can be predicted preoperatively. We specifically reviewed the effects of these endoleaks on aneurysm morphology by looking at proximal neck dilatation, aneurysm diameter, and sac volume and attempted to determine if any characteristics of collateral vessels might be associated with type-ii endoleaks. METHODS All patients treated with the AneuRx bifurcated stent-graft (Medtronic AVE, Santa Rosa, CA, USA) from October 1996 through December 1998 at Stanford University Hospital were captured prospectively in a vascular database and retrospectively reviewed. Among the 71 consecutive patients treated during this period as part of an ongoing phase-i and II multicenter clinical trials, 15 had 12-month follow-up, 5 had type-i endoleaks, 4 lacked sufficient imaging confirmation of endoleak, and 1 died within 30 days, leaving 46 patients to be analyzed for this study. Three groups of patients were identified and compared: 16 patients with persistent type-ii endoleaks (PE), 14 patients with transient (duration 6 months) type-ii endoleaks (TE), and 16 patients with no endoleak (NE). Pre- and postoperative computed tomographic angiography (CTA) and 3-dimensional (3-D) reconstructions were used to determine aneurysm morphology. Preoperative measurements included the proximal neck diameter and length and the AAA diameter and volume. Postoperatively, proximal neck dilatation and changes in aneurysm diameter or volume were determined from serial measurements during follow-up. Endoleak Assessment Before discharge, all patients underwent abdominal radiography to evaluate the position and integrity of the stent-graft, duplex ultrasound scanning, and CTA; these studies were repeated at 1, 6, and 12 months and yearly thereafter. The combination of duplex imaging and CTA allowed accurate identification of the source of any endoleak and exclusion of patients with a type-i endoleak (Fig. 1). Duplex scans were performed by a registered vascular technologist using a Sequoia 512 ultrasound system (Acuson, Mountain View, CA, USA) and a Sector V4 transducer. After the patients had fasted for 6 hours, transverse and sagittal imaging was performed to measure peak systolic diameters at the largest region of the proximal, mid, and distal segments of the abdominal aorta. Close attention was given to the stent device in gray scale and color Doppler scanning of the aneurysm sac to rule out endoleak. Duplex was used to determine presence or absence of flow in the aneurysm sac and branch vessels. Quantitative assessment of flow velocities and direction was performed for the patent branches and aneurysm sac. All scans where reviewed by a vascular surgeon. Spiral CT was performed with a Lightspeed Qxi CT scanner (General Electric Medical Systems, Milwaukee, WI, USA). Precontrast studies were obtained routinely. After preliminary timing of 15 ml of iodinated contrast bolus, 80 to 150 ml of nonionic iodinated contrast medium were injected at 4 ml/s. CT scans were acquired at pitch 6.0 with a 2.5-mm nominal section thickness throughout the entire scan. All images were reconstructed at in-

3 TYPE-II ENDOLEAKS 505 Figure 1CTA (top) and duplex ultrasound images of type-ii endoleak originating from a lumbar artery. tervals equal to 50% of nominal section thickness and viewed interactively on a workstation. Collateral Vessel Quantification IMA patency was defined as visualization of the vessel in communication with the aneurysm on preoperative CTA. The diameter was measured on the CTA images using calipers; 3 independent measurements were obtained and averaged. Similarly, all lumbar arteries visualized on the preoperative CTA from the renal arteries to the level of the aortic bifurcation were counted. Statistical Analysis Results are expressed as mean SD. Intergroup differences were compared using an analysis of variance (ANOVA) for 3 independent samples. Cox proportional hazards regression analysis was employed to identify any independent risk factors for persistent type-ii endoleaks. Kaplan-Meier estimates were used to determine freedom from aneurysm rupture and surgical conversion; p 0.05 was considered significant. RESULTS The patients in each subgroup were equally matched with regard to preoperative comorbidities and age. Mean follow-up was 21.7 months for all 46 patients; there was no difference in mean follow-up among the groups: months (range 12 36) for PE, months (range 12 37) for TE, and months (range 12 46) for NE. In terms of aneurysm morphology, there were no significant intergroup differences in the dimensions of the proximal neck (Table 1); however, the mean aneurysm diameter in the TE group was slightly larger (p 0.04). The

4 506 TYPE-II ENDOLEAKS TABLE 1 Preoperative Aneurysm Morphological Measurements From CTA Images Persistent Endoleaks (n 16) Transient Endoleaks (n 14) No Endoleak (n 16) p Proximal neck diameter (mm) Proximal neck length (mm) AAA size (mm) Patent IMAs IMA diameter (mm) Number of lumbar arteries (19 26) (10 46) (47 62) 81% Values are mean SD (range) unless otherwise noted. NS not significant, IMA internal mesenteric artery (19 25) (10 40) (46 80) 43% (16 26) (12 40) (48 68) 31% NS NS IMA was patent in 13 (81%) of 16 PE patients versus 6 (43%) of 14 TE and 5 (31%) of 16 NE patients, respectively (p 0.01). The mean IMA diameter was also significantly different (p 0.01) among the groups, being largest in the PE group ( mm) and smallest in patients without endoleaks ( mm). There were also significantly more patent lumbar arteries visualized on CTA in the PE group (p ). According to the Cox proportional hazards regression analysis, both a large (6 mm) patent IMA (RR 0.82, 95% CI, 0.71 to 0.94; p 0.01) and the number of patent lumbar arteries (RR 0.40, 95% CI, 0.26 to 0.62; p ) were independent risk factors in predicting persistent endoleaks. In follow-up, there was no significant change in the mean proximal neck diameters from baseline (Table 2, Fig. 2) in any group or any difference based on type-ii endoleak status. There was no change in aneurysm diameter from the preoperative value in the PE group ( mm), whereas the aneurysm shrank in the TE ( mm) and NE ( mm) groups (p 0.001). The same trend was noted in aneurysm volume (Fig. 3): no change in the PE group ( ml) but declines in both the TE ( ml) and NE ( ml) cohorts (p 0.05). No secondary interventions were performed in the NE or TE patients. The mean time for resolution of the endoleak in the TE group was 2.1 months (range ). Secondary interventions were performed in 5 (31%) of 16 PE cases over a mean 16.6 months (range 6 24); however, coil embolization of 4 lumbar arteries via the hypogastric artery and one IMA via the superior mesenteric artery, although technically successful, did not obliterate the endoleak. TABLE 2 Postoperative Changes in Aneurysm Morphology Versus Baseline Persistent Endoleaks (n 16) Transient Endoleaks (n 14) No Endoleak (n 16) Total (n 46) p Proximal dilatation (mm) AAA diameter (mm) AAA volume (ml) (8 to12) (40 to 114) Values are mean SD (range). NS not significant, AAA abdominal aortic aneurysm (20 to 5) (56 to 44) (20 to 1) (40 to 26) NS

5 TYPE-II ENDOLEAKS 507 Figure 2Comparison of pre- and postoperative neck diameters. Figure 3Comparison of pre- and postoperative aneurysm volumes.

6 508 TYPE-II ENDOLEAKS Figure 4Kaplan-Meier estimates for freedom from aneurysm rupture. Kaplan-Meier estimates for freedom from aneurysm rupture (Fig. 4) and surgical conversion (Fig. 5) were 100%, 100%, and 100% at 6-, 12-, and 24-months in the 46 subgroup patients. Among the entire 71-patient population, the overall 30-day mortality rate was 1.4%; there were no aneurysm-related deaths after 30 days. There were no aneurysm ruptures and one surgical conversion. Kaplan- Meier survival estimates at 1-, 2-, and 3- years were 88%, 80%, and 74%. DISCUSSION To make our investigation of type-ii endoleaks as accurate as possible, we confined our study to patients who had 12 months follow-up and no type-i endoleak, which could mask the effect of a type-ii endoleak. Moreover, we included only patients who had type- II endoleak confirmation by both duplex and CTA imaging; however, this also excluded 4 patients diagnosed with a type-ii endoleak by only one of the tests. Our results demonstrated that the presence of both a patent IMA and 2 patent lumbar arteries are independent preoperative risk factors for persistent type-ii endoleak following endovascular repair. The greater the number of patent lumbar arteries, the higher the risk. Patients with persistent endoleaks were much more likely to have a patent IMA on preoperative contrast CT scan (81%) as compared Figure 5Kaplan-Meier estimates for freedom from surgical conversion. to patients with transient endoleaks (43%) or no endoleaks (31%) (p 0.01). However, 31% of patients with no documented evidence of an endoleak on follow-up imaging had a patent IMA. We determined that not only was IMA patency important, but also the diameter of the IMA. Patients with persistent endoleaks had a significantly larger IMA than those in the other groups. While one might conclude that patients with a large, patent IMA should undergo preoperative embolization in order to decrease the risk of postoperative persistent type-ii endoleaks, the clinical benefit of such an approach has not yet been proven. Preoperative embolization of the IMA may increase the risk of colon ischemia following endovascular repair by occlusion of distal branches of the IMA, thereby obstructing collateral flow. The risk would be especially great if collateral flow to the pelvis was already compromised or the internal iliac artery was covered with the endovascular graft. Thus, a more conservative approach to see if the endoleak resolves spontaneously appears reasonable, reserving IMA embolization only if clinically indicated. This is especially true in patients with intermediate-sized IMAs and a smaller number of visualized lumbar arteries, as demonstrated in our group with transient endoleaks.

7 TYPE-II ENDOLEAKS 509 Our data showed that patent lumbar arteries may play a more significant role in the development of type-ii endoleaks than even a large, patent IMA. There were significantly more lumbar arteries visualized in patients with persistent endoleaks versus those with transient or no endoleak, respectively. In our series, retrograde flow from the lumbar arteries, easily demonstrated by both duplex and dynamic spiral CTA with 3-D reconstruction, was a primary source of type-ii endoleak. This was confirmed by selective hypogastric arteriograms indicating retrograde flow into the aneurysm sac through the lumbar arteries. Thus, preoperative embolization of the IMA only could be futile since flow would persist through the lumbar arteries. In 4 patients with persistent endoleak, selective catheterization of the hypogastric artery and coil embolization of the branch vessel was technically successful but failed to obliterate the endoleak due to IMA flow. The long-term changes observed in aneurysm morphology among our patients suggest several conclusions. First, as regards the proximal neck, there was no clinically significant neck dilatation, which could be a possible source of late proximal type-i endoleak. Even the minimal amount of dilatation measured probably was attributed to graft oversizing at the time of the original procedure because the increase in neck diameter appeared after the initial postprocedural CTA and did not change over time. More importantly, the presence of type-ii endoleak seemed to have no effect on the proximal neck and proximal stent-graft fixation. Therefore, the risk of developing a late proximal endoleak appears small in well-selected patients up to 3 years after endografting. Aneurysm diameter and volume demonstrated significant differences among the 3 groups. While the aneurysms with persistent endoleaks increased in diameter and volume as compared to the other groups, both of which decreased, the overall increase was almost negligible. Most aneurysms with persistent endoleaks remained relatively unchanged, 2 aneurysms enlarged, and one aneurysm decreased in volume. Those with transient endoleaks tended to decrease in size or remain relatively unchanged. In aneurysms with no endoleaks, more than two thirds (69%) decreased in volume; 5 were unchanged. Evaluation of clinical success includes, most importantly, prevention of aneurysm rupture, which was achieved in all our patients. Others have reported aneurysm rupture with type-ii endoleaks, 2,3,15 17 but these have always been associated with a type-i endoleak as well. We were scrupulous in eliminating coexistent type-i endoleaks in this study, and there is no evidence in our data that type-ii can lead to rupture. The small diameter enlargement (1.8 mm) in the PE group may be clinically insignificant and may not portend an increased risk of rupture. If AAA size increases 5 mm, further treatment (endovascular or open) should be carried out. When looking at broader measures of successful endovascular repair, such as freedom from aneurysm rupture and from surgical conversion, all 3 groups were equivalent (100%), so the presence of a type-ii endoleak made no significant difference on overall outcomes. Therefore, neither a large, patent IMA nor 2 lumbar arteries should exclude a patient from having an endovascular repair. Zarins et al. 6 have demonstrated that endoleak is not a reliable predictor of patients who may be at future risk of rupture; they claimed that adequate stent-graft fixation and changes in aneurysm size are more important. In our study, type-ii endoleaks had little effect on increasing aneurysm size and no effect on stent-graft fixation. In conclusion, patients at risk for type-ii endoleaks after endovascular repair can be predicted preoperatively if there is either a large, patent IMA or 2 patent lumbar arteries. Persistent type-ii endoleaks are associated with a negligible increase in AAA size, while aneurysms with transient or no endoleaks will decrease in size. The presence of a type-ii endoleak has no effect on long-term infrarenal neck diameter and stent-graft migration and is not associated with aneurysm rupture in our experience. Longer follow-up is needed to confirm these observations. REFERENCES 1. May J, White G, Waugh R, et al. Life-table analysis of primary and secondary success follow-

8 510 TYPE-II ENDOLEAKS ing endoluminal repair of abdominal aortic aneurysms: role of supplementary endovascular intervention in improving outcome. Eur J Vasc Endovasc Surg. 2000;19: Schurink GWH, Aarts NJM, van Bockel JH. Endoleak after stent-graft treatment of abdominal aortic aneurysm: a meta analysis of clinical studies. Br J Surg. 1999;86: White GH, Yu W, May J, et al. Endoleak as a complication of endoluminal grafting of abdominal aortic aneurysms: classification, incidence, diagnosis, and management. J Endovasc Surg. 1997;4: Buth J, Laheij RJF, on behalf of the EUROSTAR Collaborators. Early complications and endoleaks after endovascular abdominal aortic aneurysm repair: report of a multicenter study. J Vasc Surg. 2000;31: Chuter TAM, Faruqi RM, Sawhney R, et al. Endoleak after endovascular repair of abdominal aortic aneurysm. J Vasc Surg. 2001;34: Zarins CK, White RA, Hodgson KJ, et al. Endoleak as a predictor of outcome after endovascular aneurysm repair: AneuRx multicenter clinical trial. J Vasc Surg. 2000;32: Dias NV, Resch T, Malina M, et al. Intraoperative proximal endoleaks during AAA stent-graft repair: evaluation of risk factors and treatment with Palmaz stents. J Endovasc Ther. 2001;8: Hölzenbein TJ, Kretschmer G, Dorffmer R, et al. Endovascular management of endoleaks after transluminal infrarenal abdominal aneurysm repair. Eur J Vasc Endovasc Surg. 1998;16: Zarins CK, White RA, Schwarten D, et al. AneuRx stent graft versus open surgical repair of abdominal aortic aneurysms: multicenter prospective clinical trial. J Vasc Surg. 1999;29: Becquemin JP, Lapie V, Favre JP, et al. Midterm results of a second generation bifurcated endovascular graft for abdominal aortic aneurysm repair: the French Vanguard trial. J Vasc Surg. 1999;30: Petrik PV, Moore WS. Endoleaks following endovascular repair of abdominal aortic aneurysm: the predictive value of preoperative anatomic factors a review of 100 cases. J Vasc Surg. 2001;33: May J, White G, Yu W, et al. Concurrent comparison of endoluminal versus open repair in the treatment of abdominal aortic aneurysms: analysis of 303 patients by life table method. J Vasc Surg. 1998;27: Harris PL. The highs and lows of endovascular aneurysm repair: the first two years of the EU- ROSTAR Registry. Ann R Coll Surg England. 1999;81: Görich J, Rilinger N, Söldner J, et al. Endovascular repair of aortic aneurysms: treatment of complications. J Endovasc Surg. 1999;6: Bohm T, Soldner J, Rott A, et al. Perigraft leak of an aortic stent graft due to material fatigue. AJR Am J Roentgenol. 1999;172: Krohg-Sorensen K, Brekke M, Drolsum A, et al. Periprosthetic leak and rupture after endovascular repair of abdominal aortic aneurysm: the significance of device design for long-term results. J Vasc Surg. 1999;29: Zarins CK, White RA, Fogarty TJ. Aneurysm rupture after endovascular repair using the AneuRx stent graft. J Vasc Surg. 2000;31: