Author Affiliations:

From the ¹Hospital Clinic of Barcelona, Barcelona, Spain and 2Eurostar Data Registry Center, Catharina Hospital, Eindhoven, The Netherlands.

Correspondence: Vincent Riambau, MD, Hospital Clinic of Barcelona, Vascular Surgery, Villarroel 170, Barcelona, Spain 08036. E-mail: riambau@meditex.ese.

Manuscript submitted August 24, 2008, provisional acceptance given September 30, accepted October 20, 2008.

Disclosure: The authors report no financial relationship or conflicts of interest regarding the content therein.

_____________________________

Abstract

Background. Endograft design could influence the outcome of endovascular aneurysm repair (EVAR). Objective. To determine how an aorto-uni-iliac (AUI) endograft performs compared with bifurcated (BF) designs in a mid long-term follow-up period. Patients and Methods. We retrospectively analyzed two prospective databases. A EUROSTAR cohort of 5627 patients (5274 BF endograft and 353 AUI) and a personal series of 291 (255 BF and 36 AUI) elective EVAR cases were reviewed. In both sub-studies, old generation endografts were excluded. AUI patients were significantly older than the BF patients. Patients unfit for open repair were more frequently found in the AUI group. Distal aortic diameter was significantly smaller in AUI patients. Occlusion of common iliac arteries was predominantly found in AUI patients. Results. In the EUROSTAR study, 30-day mortality was higher in the AUI group (8.5% vs 2.8%, p < 0.0001). Annual mortality was 7.3% for BF devices and 14.1% for AUI devices (p < 0.0001). In our series, no differences were observed between both groups in terms of 30-day mortality (2.35 vs 2.77, respectively, p = 0.87). Nevertheless, after 7-year follow up, freedom from any cause mortality was lower in the AUI group (54.9% vs 82.4%, respectively, p > 0.001). No other significant differences were observed in terms of endoleak rate, graft migration, graft patency, transfemoral and abdominal reinterventions, aneurysm rupture, graft infection, and pseudoaneurysm formation in both studies. Conclusions. The performance of AUI endografts were similar to BF endografts, even in patients with poor anatomy and medical condition. Only differences in early and cumulative mortality were found. The AUI modality allows the treatment of high-risk patients with unfavorable anatomy and expands EVAR applicability. Further randomized trials should be performed to certify this equivalent behavior among these endografts designs.

Introduction

Endovascular aneurysm repair (EVAR) has been accepted as an alternative to open surgery, especially for high-risk patients. But despite promising results, the need for re-interventions and required surveillance and follow-up protocol have diminished its stance against open surgery as seen in randomized prospective clinical trials.1,2 Endograft design and configuration can influence EVAR outcome in this regard; however, the issue has not been the focus of serious study.

Almost from the very beginning, two endograft configurations emerged rapidly as most useful in the treatment of the majority of patients: the bifurcated (BF) and aortouniiliac (AUI) endografts. Similar to the situation found in open repair, the BF endograft was immediately embraced as the most desirable for EVAR because it seemed more anatomically correct and likely more durable than its aorto-aortic and AUI counterparts. Additionally, it was backed by a rather large and long surgical experience extending back to the inception of modern vascular surgery. The AUI graft has been conceptually handicapped by the required femoro-femoral bypass that is often considered to represent a liability in terms of the potential for graft-related complications such as thrombosis, infection and anastomotic pseudoaneurysms.3 While BF endografts are thus considered to be best, they are not always applicable. Extreme iliac artery tortuosity, unilateral iliac artery occlusion, and a narrowed aortic bifurcation are some of the most frequently encountered anatomical limitations that can preclude the use of a BF endograft. Another “soft” reason — some have advanced a preference for AUI over BF grafts related to the need for contralateral-stump cannulation when constructing a BF stent-graft. When used in the context of arterial occlusive arterial disease, the patency of femoro-femoral bypass grafts has proven inferior to that of aortobifemoral bypass.4 However, the patency may be higher and more acceptable when used for the treatment of aneurysmal (not occlusive) disease.5

EVAR using AUI endografts has been shown to produce acceptable results in both single center5–11 and multicenter experiences.12 In order to confirm these impressions, and to further substantiate (perhaps) the use of AUI graft configurations during EVAR, we set out to conduct a retrospective review of the available EUROSTAR database and from our own single-center experience.

Patients and Methods

A 5,627-patient cohort was reviewed from the EUROSTAR database. Old-generation endografts were excluded. Data were collected prospectively but analyzed retrospectively. Demographics and baseline characteristics of 5274 patients having a BF endograft and 353 with an AUI graft were compared (Tables 1 and 2). AUI patients were significantly older. Patients unfit for open repair were more frequent in the AUI group (33.0% vs. 21.8%, p < 0.0001). Bigger aneurysms were more frequently treated with AUI (58.5 ± 10.7 vs. 61.6 ± 14.6, BF vs. AUI, p < 0.0022). The distal aortic diameter was significantly smaller in AUI patients (BF=25.9 mm vs. AUI=29.8 mm, p < 0.0001). Occlusion of a common iliac artery was predominantly found in AUI patients.

From our own series, we reviewed elective EVAR patients treated from 1997 to 2005 when 36 patients had EVAR with an AUI device and 255 had BF endografts. During AUI procedures, the femoro-femoral bypass was constructed with an 8mm externally supported Dacron vascular graft (Fluoropassive, Vascutek Inc, Scotland, UK). The mean age was similar in both groups (76.70 years for BF and 79.61 for AUI). As in the EUROSTAR registry, patients unfit for open repair tended to be treated with AUI rather than a BF graft (39.7% vs. 24.7%, p < 0.0001). Predictably, ASA classification III-IV was more frequent on AUI patients (94.4% vs. 85.5%, p = 0.010). Neck diameter was not significantly different in the two groups (26.0 mm vs. 24.2 mm, p = 0.186). However, mean AAA diameter was larger on AUI patients (63.1 mm vs. 58.9 mm, p = 0.054). The follow-up protocol was similar in both series and in adherence to EUROSTAR recommendations.13 In the EUROSTAR groups, the mean length of follow-up for BF endografts was 24 ± 18 months (range 0–96), and for AUI configurations was 20.7 ± 18.9 months (range 0–96). In our own series, the mean length of follow-up for BF endografts was 18.8 months (range 1–96 months), and it was 17.3 months (range 1–84) for AUI patients. Statistical methodology involved Chi Square for discrete and Mann-Whitney for continuous variables. The long-rank test was applied for late result analysis. SPSS software (SPSS Inc. Chicago, Illinois) was used to perform the statistical analysis. P values of 0.05 or less were considered statistically significant.

Results

EUROSTAR patients. The 30-day mortality was higher in the AUI group (8.5% vs. 2.8%, p < 0.0001). Mortality per year was 7.3% for BF devices and 14.1% for AUI devices (p < 0.0001) (Figure 1). No other significant differences were observed in terms of endoleak rate, graft migration, graft patency (Figure 2), aneurysm rupture, graft infection, or pseudoaneurysm formation. However, there was a trend where AUI endografts had an apparent advantage in several aspects such as: lower rate of migration (3.4% vs. 8.3%), lower re-intervention rates (10.7% vs. 17.6%), less likely to undergo conversion to open repair (4.1% vs. 7.6%), and a lower rate of aneurysm rupture (1.0% vs. 2.5%). All-cause mortality, however, was significantly higher for AUI patients (39.5 vs. 30.8%, p < 0.0001). Aneurysm-related mortality was also significantly higher in the AUI group (11.3% v.s 11.3%, p < 0.0001). In our own series, there was no difference in 30-day mortality between the two groups (2.3 vs. 2.7%, p = 0.87), but after 7 years (Figure 3), the all-cause mortality trend favored the BF group (54.9 vs. 82.4%, p > 0.001).

The results were also scrutinized for other complications. No graft infections occurred in AUI patients and 1 in the BF group. Pseudo-aneurysm formation was not observed in any case. Iliac limb thrombosis occurred in 2 cases in the BF group, and 1 AUI patient had aortic endograft thrombosis associated with a graft kink. The latter was treated with axillo-femoral bypass. Six BF patients were converted to AUI for treatment of graft migration. AAA rupture occurred in 2 BF patients and none in the AUI group. There were no significant differences with any such complications between the AUI and BF patient groups.

Looking for any kind of complication, except mortality, BF endografts included 26% of the adverse events during 7-year follow up and AUI configurations registered 35% (p = 0.10). In spite of the fact that overall mortality looks higher in the AUI group (Figure 3), aneurysm-related mortality was similar for both groups after 7-year surveillance (2.4%, BF vs. 2.7%, AUI).

Discussion

These data suggest at least that AUI performs similar than BF in terms of durability and complication rates, even with potential worse anatomies. But the more severe medical conditions increase the early and late mortality rates in AUI patients. Finally, there is no reason to maintain previous concerns about infection, patency, and pseudoaneurysm formation, especially related to AUI endografts.

The most important message to emerge from this review is that the AUI graft would seem to emerge as a reasonable and good alternative to BF endografts in the treatment of AAA. It was especially comforting to see that AUI endografts were not accompanied by an increased incidence of thrombosis, infection, pseudo-aneurysm or re-intervention, precisely the kind of reasons often voiced to justify relegating AUI options to a secondary role for management of most AAA patients. In fact, some of these problems were found to be more likely in the BF patient cohort, and some of these required conversion to AUI. However, the observation that all-cause mortality was worse in the AUI cohort could be a cause for concern, but likely related to patient selection as AUI grafts were usually used on patients with worse anatomies and overall poorer state of health.

Historically, the AUI endograft was adopted early in the evolution of endovascular aneurysm repair because of the ability to load the stent graft onto a smaller sheath, and it proved easier to assemble and implant. This was particularly true in the days when homemade endografts first emerged.6 When the AUI system was first introduced, concerns were raised about durability of the extra-anatomic femoro-femoral bypass. It was believed that the additional graft would add morbidity to the procedure and adversely affect long-term outcome. It is generally accepted that, in the context of arterial occlusive disease, femoro-femoral bypasses produce inferior patency rates than aortobifemoral reconstructions,4 although some authors have reported exceptions to this rule.13 Early reports of femoro-femoral bypass grafting in aneurysmal disease were encouraging and suggested that, unlike the occlusive disease scenarion, it may be associated with satisfactory patency rates and low morbidity.5,14 In the Eurostar study, a 97.8% patency rate was achieved after 7-years, and it was 100% in our own series. These figures compare favorably with some of the best outcomes reported for any such reconstruction.3,12 Inflow and outflow considerations are crucial to achieve long-term patency.15 Patients with unsupported endograft limbs could be more prone to developing endograft stenosis or thrombosis.12 Fully stented endograft limbs appear to offer improved patency over unsupported limbs.

Another concern related to femoro-femoral bypass as an extra-anatomic reconstruction is the potential graft infection. In the Eurostar study, this complication was reported (0.4 % vs 0.3%, BF vs AUI, NS). In our experience, only one BF graft infection was recorded. However, from the literature, some studies reported around 2% of graft infection rate in AUI designs.11,12 The potential for direct contamination and infection of the prosthetic graft is related to duration of operative procedure, and is likely to be greater with AUI and femoro-femoral bypass than with BF devices.

In conclusion, the AUI endograft associated with a femoro-femoral bypass for EVAR in elective patients demonstrated similar outcomes to BF endograft configurations over the long term in long-term. It represents, clearly, a good endovascular strategy alternative that may be especially useful in patients with difficult or unfavorable anatomy, thereby expanding the reach of endovascular treatment.