Endovascular aneurysm repair (EVAR) has been increasingly used to treat abdominal aortic aneurysms (AAAs). Endografts have been improved to reduce complications, including endoleaks, and overcome anatomical limitations. Anatomical fixation on the aortic bifurcation and a modular proximal cuff for the aortic seal distinguishes the AFX Endovascular AAA System (Endologix) from many other proximal fixed endograft products [1]. Although the AFX with Strata, an early AFX model, has several advantages associated with anatomical fixation, there have been some reports of an increase in late type IIIa endoleaks (T3aE) [2,3]. In response, Endologix improved the device material and released the AFX with Duraply in 2014 and the AFX2 in 2016. Additionally, recommendations regarding component overlap were updated to decrease the risk of endoleaks [4].

Several studies have demonstrated a decrease in T3aE rates with the AFX and AFX2 [5]. However, whether these implants are safe with respect to T3aE remains controversial [6], particularly with regard to long-term outcomes. Here we describe a case of an AFX2-implanted AAA in which the overlap between the components was lost during follow-up due to conformational changes, resulting in a massive T3aE. This case report was approved by the Institutional Review Board of Seoul National University Hospital (no. 2102-107-1197), which waived the requirement for informed consent owing to the retrospective nature of the study.

A 67-year-old male was referred to the vascular department with aneurysmal sac enlargement 36 months after EVAR and admitted for the treatment of acute kidney injury.

The initial EVAR was performed at an outside hospital using an AFX2 endograft. The patient was on aspirin due to cerebral infarction, had a history of tumorectomy for meningioma, and was on hemodialysis for end-stage renal disease. At the time of EVAR, the aneurysm was 5.7 cm, and the infrarenal neck angle was 78°. The previous EVAR involved the AFX2 main body (diameter: 28 mm), AFX2 proximal extension (diameter: 34 mm) and a 16-13-mm Endurant II (Medtronic) limb extension after right internal iliac artery embolization (Fig. 1).

Figure 1. (A) The procedural schematic illustration and (B) a post-endovascular aneurysm repair (EVAR) fluoroscopic image. The overlap length was approximately 5 cm at the time of EVAR.

On the first follow-up computed tomography (CT) performed 12 months post-EVAR, the maximal diameter of the aneurysmal sac was 5.1 cm (Fig. 2A). However, during further surveillance on CT angiography, the sac size gradually increased to 5.6 cm at 22 months and 5.9 cm at 39 months post-EVAR (Fig. 2B). No endoleak was found in either the arterial or delayed phases of the follow-up CT angiography. To evaluate the cause of the sac enlargement, aortography was performed at 36 months post-EVAR; however, no definitive endoleak was detected (Fig. 2C). Although surgical intervention was recommended because of a possible type V endoleak, the patient declined treatment and was discharged.

Figure 2. Follow-up computed tomography and diagnostic aortography images. (A) At 12 months post-endovascular aneurysm repair (EVAR), the aneurysmal sac measured 51 mm×68 mm. (B) At 39 months post-EVAR, the aneurysmal sac increased to 59 mm×80 mm. (C) Aortography revealed no evidence of endoleaks.

At 52 months post-EVAR, the patient presented to the emergency room with anemia that developed after hemodialysis, and a CT scan revealed an enlarged aneurysmal sac measuring 6.7 cm. Component overlap had been completely lost, resulting in a large endoleak (Fig. 3A). Aortography confirmed component separation and contrast leakage, indicating the presence of a T3aE (Fig. 3B). A review of previous CT scans revealed a gradual, persistent decrease in component overlap (Fig. 4). Considering the hostile anatomy and severe neck angulation, an open surgical repair was performed. Total endograft explantation and endoaneurysmal aorto-biiliac interposition grafting were performed using a transabdominal approach. Inflow was controlled by clamping the proximal infrarenal aorta. All endografts were explanted except for the right distal endograft (Endurant limb extension), which was partially removed by transection at the level of the external iliac artery and sutured together with the remaining vessel wall at the graft-artery anastomosis. The patient was discharged 13 days postoperatively and remained well during 10 weeks of follow-up.

Figure 3. Computed tomography (CT) and aortography at 52 months post-endovascular aneurysm repair. (A) CT scan demonstrated an enlarged aneurysmal sac and contrast leakage out of the endograft (arrows). (B) Aortography revealed a loss of component overlap and contrast leakage (arrow).

Figure 4. A gradual decrease was noted in component overlap. At the time of endovascular aneurysm repair (EVAR), the overlap length was about 5 cm. A serial images of the computed tomography scans revealed progressive loss of component overlap.

Despite improvements made to the AFX2 device to address endoleak issues, controversy persists regarding its safety [7-9]. Previous studies have reported the occurrence of delayed T3aE, especially with early AFX models [6,8]. T3aE is a serious concern, as it re-exposes the AAA sac to direct arterial pressure, potentially leading to sac enlargement, rupture, and death. To prevent T3aE, preoperative planning and appropriate selection of endograft components are essential [10]. Various factors influence the long-term junctional durability of modular endograft systems, including AAA size, proper device sizing and selection, angulation at the overlap zones, maximizing overlap between modular components, usage of additional overlapping stent grafts in hostile anatomy, and avoiding excessive oversizing of the proximal extension relative to the main body [11,12].

To mitigate the endoleak issue, Endologix’s instruction for use (IFU) recommends sufficient overlap consisting of a length greater than the aneurysmal radius plus 20 mm between the distal bifurcated and proximal endograft components of its AFX2 device [4]. In the current case, with an aneurysm diameter of 57.3 mm, an ideal overlap of ≥48 mm was required, which was almost satisfied at the time of the EVAR procedure. Regarding neck angle, Endologix recommends using the AFX2 device in patients with an angle ≤60°. Our case was off-IFU in this regard, as the initial neck angle at the time of EVAR was 78°. Acute angulation in the overlap zone may have contributed to the development of the late endoleaks. While on-IFU use of EVAR devices is recommended, in inevitable cases of severe angulation, such as off-IFU, a telescopic technique during deployment has been proposed to achieve better graft–aorta attachment [13].

Although preventing endoleaks is crucial, patients with off-IFU application require more careful and prolonged surveillance. In our case, the sac initially showed an increased size without a clear endoleak and was thus diagnosed as endotension [14]. Endotension is generally attributed to unrecognized endoleaks, pressure transmission through the endograft wall, or transudate migration through the endograft owing to porosity [15]. In the case of true endotension without an undetected endoleak, blood is absent in the aneurysmal sac and rupture is unlikely to cause massive hemorrhage [16]. However, as demonstrated in our case, endotension can cause morphological changes in the aneurysm, disrupting the device and inducing a potential new endoleak. Thus, treatment is generally recommended if the sac continues to increase during surveillance, even if no evidence of an endoleak is observed.

Upon the diagnosis of the massive T3aE, open surgical treatment was chosen over endovascular re-lining. The choice of endovascular or open treatment for T3aE is complex and depends on the location of the endoleak and anatomy of the aortic neck [17]. While a recent report indicated that re-lining has a comparable aneurysm exclusion rate in failed AFX with faster recovery and lower mortality [18], the long-term outcomes and potential for re-lining to replace surgical repair in patients with hostile anatomy remain unclear. Furthermore, the AFX2 endograft has an endoskeleton design with the stent material located inside the fabric, which is fixed only at the top and bottom [19]. Such morphological characteristics may limit the ability to completely seal leakage upon re-lining with other endografts with exoskeleton designs. In our case, severe angulation of the previous AFX2 was another reason we chose surgical repair over re-lining. Safely advancing the guidewires and endografts into the angled uncoupled AFX2 with endoskeletons would have been technically challenging.

In conclusion, patients who undergo EVAR using an AFX2 require continuous surveillance and increased clinician awareness of aneurysmal sac enlargement. Despite the upgrade in AFX2, the issue of T3aE seems to persist. Especially when patients with a hostile anatomy are considered for treatment with an AFX2, ample component overlap may be necessary to prevent late T3aE. Most importantly, these patients should be carefully monitored for any conformational changes and, in cases of endoleak with sac enlargement, open repair or endovascular re-lining should be considered.

None.

The authors have nothing to disclose.

Concept and design: AH. Analysis and interpretation: MK. Data collection: MK. Writing the article: MK. Critical revision of the article: SA, SKM, AH. Final approval of the article: all authors. Overall responsibility: AH.