Successful Off-Label Use of an Iliac Branch Device to Rescue an Occluded Aortofemoral Bypass Graft

Introduction

Iliac branch devices represent a significant improvement in the treatment of common iliac artery (CIA) and aortoiliac aneurysms. They have been employed with high technical success, excellent patency, and low reintervention rates at short- and midterm follow-up.^1–3 However, with the confidence gained over the past few years using these devices electively, some authors have already described novel and off-label approaches using these devices.^4,5

We present a case in which the Gore Excluder Iliac Branch Endoprosthesis (IBE; W.L. Gore & Associates, Flagstaff, AZ, USA) was employed in an off-label fashion to rescue an acutely occluded aortofemoral bypass.

Case Report

A 52-year-old man with past medical history of hypertension, dyslipidemia, and smoking was admitted to the emergency department with acute right limb ischemia (grade 2A: sensory loss, rest pain, and mild muscle weakness). Two years before he had undergone at another hospital an aortoiliac kissing stent procedure [self-expanding Luminexx stents (Bard Peripheral Vascular, Tempe, AZ, USA)] and primary covered balloon-expandable stenting (Fluency, Bard Peripheral Vascular) of the right external iliac artery (EIA) for severe intermittent claudication (Rutherford category 3).

Physical examination revealed the absence of the right femoral pulse (present on the left side), and duplex ultrasound confirmed the thrombosis of the previously placed right EIA stent. Computed tomography angiography (CTA) documented a healthy infrarenal aorta and patent aortoiliac kissing stents, as well as both CIAs and internal iliac arteries (IIA). On the right side, the EIA covered stent, which was not overlapped with the CIA stent, was completely occluded, as were the EIA and common femoral artery (CFA), while the profunda femoris artery (PFA) was reperfused by ipsilateral pelvic collaterals. The right superficial femoral artery presented a long chronic total occlusion; the popliteal artery was reperfused in the infra-articular segment from PFA collaterals. Two tibial arteries ran down to the foot. Records of the kissing stent procedure suggested that the stents had been overlapped about 1 cm, so the CTA findings were interpreted as a displacement of the stents on the right side.

Given the immediate threat to the limb, the patient was urgently taken to the operating room and placed under general anesthesia; surgical cutdown of the right groin was performed. Intraoperatively, a dissection not evident on the preoperative CTA was seen extending deep into the PFA. CFA endarterectomy and profundaplasty were performed with 8-mm Dacron patch closure. It was thought that the displacement was caused by insufficient overlapping at the primary intervention and that the displacement itself led to the dissection.

An attempt was made to recanalize the right EIA covered stent with a 0.035-inch Terumo guidewire (Terumo, Tokyo, Japan), but it was hindered by the aforementioned displacement at the junction between the CIA and EIA. Thus, after median xypho-pubic laparotomy, a right aortofemoral bypass was performed with a 9-mm Dacron straight surgical graft. The proximal anastomosis was constructed end-to-side in the proximal infrarenal aorta (which was cross-clamped using 2 straight DeBakey clamps), while the distal anastomosis was constructed end-to-side on the femoral patch. At the end of operation, the distal anastomosis was evaluated with duplex, while the runoff to the foot was interrogated with continuous wave Doppler. No evidence of any blood flow disturbance was noted; the ankle-brachial index (ABI) on the right side was 0.8.

On the third postoperative day, the patient developed recurrent acute ischemia of the right limb. CTA showed occlusion of the right aortofemoral bypass and severe thrombus burden within the infrarenal aorta causing marked stenosis (Figure 1A-C). The aortoiliac kissing stents were patent on both sides, as were both IIAs. It was thought that plaque fractures at the site of vascular clamping during the previous open surgery were responsible for distal embolization leading to bypass occlusion.

OPEN IN VIEWER

Figure 1.

(A, B) Preoperative computed tomography angiography (CTA) coronal images documented aortic thrombus at the site of previous clamp placement, causing narrowing of the lumen. (C) CTA reconstruction of the occluded right aortofemoral bypass (upper white arrow) and right external iliac stent (lower blue arrow). (D) Intraoperative angiography confirmed the narrowing of the infrarenal aorta. (E) Fluoroscopy of the kissing balloon dilation of the iliac branch and Viabahn endoprostheses.

The CTA images were carefully reviewed; given that the proximal bypass anastomosis had been constructed in the mid segment of the infrarenal aorta, there was a sufficient healthy infrarenal neck to allow bailout endovascular repair. After discussion of all therapeutic options with the patient, an endovascular revision was elected using an in-stock Gore Excluder IBE in an off-label fashion. The patient was taken to the operating room the same day; under general anesthesia, surgical cutdowns were made in the right groin and over the left axillary artery. A straight 12-F sheath (Flexor; Cook Medical, Bloomington, IN, USA) was placed through the axillary access into the descending thoracic aorta to minimize the number of times the left vertebral artery origin was crossed during the procedure.

Intraoperatively, atherosclerotic debris was retrieved from the PFA, a finding that further confirmed the initial hypothesis. After crossing the bypass with a 0.035-inch hydrophilic Terumo guidewire, fluoroscopy-guided thrombectomy with a 5-F Fogarty catheter was performed. Subsequently, an 18-F DrySeal introducer sheath (W.L. Gore & Associates) was positioned, and a 5-F diagnostic pigtail catheter was placed in the infrarenal aorta for diagnostic angiography, which confirmed thrombus as the cause of the severe aortic narrowing (Figure 1D).

After change for a 0.035-inch stiff Amplatz Super Stiff guidewire (Boston Scientific Corporation, Marlborough, MA, USA), the main body of the Gore Excluder IBE was inserted. The initial step was to advance the device slightly beyond the level of the renal arteries, where it could be partially opened to free up the gate without engaging the surgical anastomosis. The device was then pulled back just below the left renal artery (LRA) origin, where it was completely deployed, with the EIA limb overlapping the surgical prosthesis by about 5 cm and the IIA limb opening within the aortic lumen. The IBE gate was catheterized from above using a 5-F catheter (Multipurpose; Cordis Corporation, a Cardinal Health company, Milpitas, CA, USA) and a 0.035-inch hydrophilic guidewire through a precurved 8-F Flexor sheath positioned coaxially in the 12-F sheath. The 5-F catheter and 8-F sheath were then removed, and a 13×50-mm Viabahn endoprosthesis (W.L. Gore & Associates) was inserted bare as the bridging endograft and deployed so that it landed just above the previously placed kissing stents without overlapping them. Finally, all the stent-grafts were simultaneously dilated in a kissing fashion (Figure 1E) with appropriately sized noncompliant balloons (Ultraverse; Bard Peripheral Vascular). In particular, the Viabahn endoprosthesis was gently flared in its distal third to achieve perfect apposition to the aortic wall. Completion angiography showed patency of both renal and IIA arteries, correct placement of all deployed endografts with excellent conformability, and complete recanalization of the right aortofemoral bypass.

The postoperative course was uneventful. The patient was placed on daily aspirin (100 mg) plus clopidogrel (75 mg) for the first 6 months, with lifelong daily aspirin thereafter. The patient was discharged home 3 days after the rescue procedure with a right ABI of 0.8. Clinical and radiologic follow-up visits with duplex and CTA were conducted at 1, 6, and 12 months. All imaging examinations showed satisfactory placement of the endografts, no evidence of any endograft-related complication (ie, stent fracture, thrombosis. or displacement), and patency of both IIAs as well of the right aortofemoral bypass (Figure 2). A 1 year, the patient was doing well with stable symptoms of mild intermittent claudication (Rutherford category 1) and a right ABI of 0.8.

OPEN IN VIEWER

Figure 2.

At 1 year, (A-D) sequential axial computed tomography (CT) images at 5-mm intervals showed the interaction between the Viabahn endoprosthesis and the iliac kissing stents. The distal radiopaque markers of the Viabahn endoprosthesis are landed upon the iliac kissing stents without overlapping them. (E, F) CT reconstructions demonstrated regular placement of all deployed endografts: the iliac branched endoprosthesis is at the infrarenal position and the Viabahn stent-graft is well conformed within the infrarenal aorta. The internal iliac artery is patent on both sides.

Discussion

Endovascular surgery has been described as a safe and feasible approach for repair of recurring aortic lesions after open surgery.^6,7 Indeed, it is well known that open reoperations are often more technically demanding than the primary surgical procedure. Given the potential role for rescue of failed open aortic procedures by means of endovascular techniques, we believe that, when technically feasible, a surgical operation should be performed to allow straightforward secondary endovascular repair should the need arise. In this case, the Gore Excluder IBE was used as a bailout strategy for management of an early failure of an aortofemoral graft. The main considerations in this situation (eg, potential disruption of a fresh aortic anastomosis) could be very different as compared with a late rescue. However, given that the endograft was overlapped with the previous surgical graft for a long segment, it was assumed that this would ensure adequate protection from any issue occurring at the level of the aortic anastomosis.

Other endovascular alternatives would have been possible in our case, including kissing stents or the covered endovascular reconstruction of the aortic bifurcation (CERAB) technique. ⁸ We believe that using a unibody stent-graft may offer several potential advantages compared with these endovascular alternatives, such as minimizing potential lumen narrowing and flow disturbances introduced by multiple covered stents, trapping the atherosclerotic material that could potentially embolize, and making future “up-and-over” interventions to the lower extremities less technically challenging. Indeed, this is complemented by in vitro models supporting the concept that a more anatomic and physiologic endovascular reconstruction entails less hemodynamic disturbance, which may in turn lead to increased stent patency both in the short and long term.^8,9

There are some technical issues from our case that are worthy of discussion. First, we adopted an antegrade axillary access, which had proven feasible for catheterization of the Gore Excluder IBE to reduce the risk of iliac stent manipulation and displacement in our previous experience. ¹⁰ Moreover, it would have provided easy access to both kissing stents should additional maneuver on one side have been required. The axillary artery access was performed surgically because this approach is safer and more expeditious than high brachial artery access when ⩾12-F devices are used.

There was a relatively short distance from the origin of the LRA to the proximal anastomosis of the bypass, which might have caused incomplete opening of the IIA limb of the device. However, the Gore Excluder IBE provides very accurate positioning through its repositionable delivery system, and the option for repositioning allowed us to partially open the device above the LRA without any difficulty pulling it back to the final infrarenal position. Furthermore, the flexibility of the IBE main body has proven to be highly conformable with the anatomy, providing kink resistance even in case of severe tortuosity and reducing the risk of limb thrombosis related to anatomic constraints. ¹¹

One possible shortcoming of using the Gore Excluder IBE without proximal fixation from a conventional stent-graft might relate to the risk of late migration since recommendations from the manufacturer state that a proximal aortic endograft in the infrarenal aorta is mandatory to guarantee proper proximal sealing of the IBE. However, Giaquinta et al ¹ did not encounter migration in their 49 CIA aneurysms treated solely with an iliac branch device. Similar results were found in the pELVIS Registry, ¹² which showed that when a proper anatomic situation is provided, single iliac branch device placement may be as safe and effective as more extensive procedures. Indeed, our patient had a healthy proximal neck in the first 15 mm below the LRA. Thus, the ⩾20% oversize provided at this level should provide sufficient radial force to avoid late failure. Furthermore, a major part of the endograft was positioned within the existing surgical graft, resulting in a double layer configuration that should be more durable than a primary endovascular repair. Nevertheless, considering the lack of long-term data for this application, lifelong follow-up is necessary to assess device integrity over the long term.

We sought to preserve flow to both IIAs since CTA findings at admission showed reperfusion of the right PFA from ipsilateral pelvic collaterals. The choice not to use a bare metal stent landing directly within the left iliac stent was taken because, in our opinion, this would have left the aortic thrombus uncovered, with subsequent risk of further embolization. Subsequently, a covered stent was placed to exclude the mural thrombus from the circulation and deployed so that it landed just above the edge of the existing iliac stents.

The IIA component of the Gore Excluder IBE was not used since its length would have made it mandatory to overlap it with one of the aortoiliac kissing stents. Thus, the Viabahn endoprosthesis was used as the bridging endograft because of its favorable characteristics, which include flexibility, fracture resistance, and a low restenosis rate. In addition, it has a laser-cut contoured proximal edge, which may improve apposition to the vessel wall, and its successful use in this application is well supported by the current literature. ¹³ In our opinion, even if the device used did not offer significant oversizing in the aorta, it would not be an issue since there was no aneurysm.

Based on other experiences,^2,3 two technical aspects should be underlined in order to avoid occlusion of the Gore Excluder IBE. The first is to avoid extending the bridging endograft too high, with its proximal end above the gate of the main body. This is concordant with the notion that stent protrusion in the kissing configuration creates radial mismatch (gutters around the stents) and protrusion mismatch (the overlap distance of the stents) that could negatively impact patency. ¹⁴ The second is to ensure adequate runoff distal to the endovascular reconstruction since there is evidence that poor outflow is a significant risk factor for stent failure. ¹⁵

Conclusion

The Gore Excluder IBE may represent a versatile solution for the rescue of cases when open surgery would be associated with a considerable risk. This off-label application of a well-recognized endovascular device appeared safe and feasible and may prove useful as a valuable alternative in properly selected patients. Familiarity using the device within the instructions for use is recommended before attempting off-label complex cases such as described here.