Physician-modified fenestrated endovascular repair of type 1A endoleaks from polymer-based low-profile endografts

Introduction

Endovascular aortic aneurysm repair (EVAR) has replaced open repair as the dominant treatment modality for abdominal aortic aneurysm (AAA). ¹ , ² With accumulating experience and refinement of technology, EVAR has expanded to include large, short, and angulated infrarenal necks, collectively termed “hostile necks.” ³ Newer generation devices have been introduced with design features aiming to provide improved seal in hostile necks. One such device is the Ovation system (Endologix, Irvine, CA), which is designed to achieve proximal seal using a polymer-based technology (O-ring), while the fixation is provided by two layers of proximal bare metal stents (Figure 1). The unique design features of this device result in large double-layered suprarenal fixation stents and a short mainbody above the flow divider. The unique features of the Ovation system pose several challenges for rescue of type 1a endoleaks with fenestrated endovascular aortic repair (FEVAR). We describe two such cases and the technical tips used to overcome these challenges. Both patients provided their consent for the off-label procedures as well as publication of their cases. Institutional Review Board approval was waived.

OPEN IN VIEWER

Figure 1.

The Ovation system is demonstrated (a). Proximal seal is supported by its unique polymer-based technology and two layers of proximal fixation bare metal stents (b) while iliac limbs accommodate a variety of iliac artery sizes (c). Images obtained courtesy of Endologix.

Case 1

A 93-year-old man underwent an EVAR with Ovation at outside hospital, three years prior to presentation. Due to stage 3 chronic kidney disease, he was undergoing surveillance with non-contrast CT scans. The aneurysm sac showed progressive expansion and was urgently referred to us when he developed flank pain. Computer Tomography Angiography (CTA) demonstrated a type 1a endoleak with an 8.8 cm AAA (Figure 2(a) and (b)). This was repaired with six vessel (common hepatic, splenic, superior mesenteric, two right renal, left renal arteries) physician-modified fenestrated cuff, using a tapered Zenith Alpha stent graft (Cook Medical, Bloomington, IN). The modification technique included placement of preloaded wires, as previously described. ⁹ Under general anesthesia, a left high brachial cut down was performed for branch vessel cannulation, and percutaneous femoral access was used for aortic component implantation. Six branch vessels were sequentially catheterized with staggered deployment of the fenestrated endograft (Figure 2(c) and (d)). Fluoroscopy time was 95 min and total contrast use was 120 mL. After uneventful hospital course, he presented one month later with claudication symptoms from occlusion of the main body graft above the flow divider, which was revised with deployment of Viabahn VBX stent grafts (W.L. Gore and Associates, Flagstaff, AZ) within the main body. The patient is doing well with no aneurysm sac growth or endoleak, 15 months after the fenestrated repair.

OPEN IN VIEWER

Figure 2.

In case 1, Axial (a) and coronal (b) CTA demonstrated robust type 1A endoleak that lead to sac expansion delineated by the solid line. Physician-modified six-vessel fenestrated cuff was deployed with the assistance of preloaded wires (c). Completion angiogram demonstrated aneurysm sac exclusion and branch preservation (d).

Case 2

An 81-year-old man underwent an EVAR with Ovation at outside hospital, seven months prior to presentation. He had a type 1a endoleak after the initial Ovation deployment despite repeated proximal balloon molding. Subsequently, he underwent an attempt at proximal extension with renal snorkels, but this was aborted due to failure to catheterize the renal arteries. Instead, a Palmaz balloon expandable stent (Cordis, Santa Clara, CA) was deployed across the renal origins, but the type 1a endoleak persisted. Within two months, he became symptomatic with progressive expansion of his AAA and transferred to our center (Figure 3(a) and (b)). At our center, pre-stenting of the renal arteries (Figure 3(c)) was performed, prior to a four-vessel physician-modified fenestrated cuff placement (Figure 3(d)). In order to prevent occlusion due to infolding of the fenestrated cuff within Ovation main body, the flow divider was raised with kissing 8 L VBX stents from bilateral groins (Figure 3(d)). Fluoroscopy time was 80 min and contrast use was 120 mL. He recovered well without limb issues and is doing well with no aneurysm sac growth six months after fenestrated repair (Figure 3(e)).

OPEN IN VIEWER

Figure 3.

In case 2, axial (a) and sagittal (b) CTA demonstrates prominent type 1A endoleak referenced by the dashed line. Palmaz stent deployment did not achieve endoleak resolution (c). Pre-stenting of bilateral renal arteries was completed at our center (c) prior to four-vessel physician-modified fenestrated cuff placement in addition to kissing iliac stent deployment (solid arrows) across the main body of the Ovation main body (d). Four-month follow-up CTA demonstrated stable aneurysm sac and branch stent patency.

Discussion

While hostile necks with the conventional EVAR devices have been associated with increased risk of adverse outcomes such as type 1a endoleak, sac expansion, and even mortality, newer generation devices designed to improve hostile neck sealing have been introduced with commercial approval for shorter necks.^4–6 These novel design features detailed above (Figure 1) and the engineering principles may represent significant improvement from the conventional endograft technology.

Current instructions for use (IFU) of the Ovation system do not have a recommended minimum length of the infrarenal neck. Rather, it specifies the range of aortic diameter (16–30 mm) at the location of the O-ring (13 mm below the lowest renal artery). This is because the polymer-based seal does not depend on a length of parallel aortic wall where the endograft is opposed. Contrary to the self-expanding stent graft, the point of circumferential contact by the polymer ring does not have to be more than the height of the polymer ring. This property of the Ovation system provides an opportunity to seal in hostile infrarenal necks as long as a contact point of the O-ring falls within the diameter, irrespective of the rest of the neck length and morphology. However, extremely precise deployment is required in order to achieve secure seal in hostile necks. This is because, if misplaced the O-ring may fail to seal, resulting in proximal failure. In other words, while the infrarenal neck in the Ovation system does not provide additional sealing, it does provide room for deployment error. Furthermore, it is worth noting that the excellent results achieved in their pivotal trial were in patients with the mean infrarenal neck length of 23 mm. ⁷ As our cases illustrate, application of this device in the hostile neck can still lead to proximal failures due to non-device-related factors such as aneurysmal disease progression and deployment error. Finally, polymer leaks during deployment, which led to recall of the Ovation system, could lead to decreased filling of O-ring and proximal sealing issues. For this reason, we caution against wide application of the Ovation system in hostile necks without overcoming the initial learning curve in precise deployment. Recently, the next generation polymer-based proximal sealing EVAR device has been launched with the design modification to allow even shorter neck sealing. However, we suggest that the same caution should be practiced in the Alto device when faced with hostile neck.

When applying proximal extension with fenestrated stent graft, we encountered specific challenges, as a result of the design of the Ovation system. First, the presence of double-layered suprarenal fixation stents can make branch cannulation more challenging. Furthermore, use of Palmaz stent can make branch cannulation even more difficult, if not impossible. In these situations, pre-cannulation of the target vessels can be helpful because it demonstrates feasibility of target vessel incorporation. Pre-cannulation of target vessels also provides the opportunity to pre-stent them, in order to facilitate FEVAR. Second, the short length of the mainbody, as well as the reinforced with polymer ring design, results in severe mismatch between the diameters of the proximal and distal seal zones of the fenestrated cuff. In both of our cases, the smallest luminal diameter across the O-Ring was 11 mm. The fenestrated devices used in these cases were 26 mm and 30 mm distally, resulting in 160% and 200% oversizing at that segment, respectively. This can result in significant infolding of the fenestrated cuff, which then can compromise both distal seal and iliac limb patency. In the first case, the mainbody occlusion was addressed with a single 8 mm × 39 mm Viabhan VBX (W.L. Gore & Associates, Flagstaff, AZ) was deployed across the occlusion and post-dilated to 14 mm. Alternatively, infolding can be addressed by prophylactic deployment of balloon expandable covered stents to create a pleating effect at the time of FEVAR. Third, staggered deployment of the fenestrated aortic component facilitates target vessel cannulation in challenging anatomy created by the presence of previous endografts, while preserving working room for branch cannulation.

Open conversion with partial and full explant is an alternative treatment option; however, both of our patients were high risk for open repair. Particularly for the Ovation system, the presence of the suprarenal bare metal stents which often extend to the level of the celiac origin would require supraceliac clamping. Complex open conversion after failed EVAR has shown a 30-day morbidity and mortality up to 40% and 13%, respectively. ⁸

Use of physician-modified endograft (PMEG) technique in this setting, as an off-label procedure is fundamentally different from off-label application of EVAR devices in short neck. While both procedures are off-label and some may view alterations of the commercially available endografts as a higher risk endeavor, correctly performed PMEG can provide an endograft that is customized to the patient anatomy and extends the proximal seal zone to a healthy aortic segment. We have previously reported the technique of adding preloaded wires to modified Zenith Alpha thoracic stent graft. ⁹ In both patients, full disclosure of off-label nature of PMEG devices as well as discussions regarding alternative treatment options including transfer to another center with investigational device exemption (IDE) protocol were performed. Based on the results of off-label PMEG operations including the two cases presented in this article, the senior author (SH) has obtained the Food and Drug Administration (FDA) approval for a Physician-Sponsored Investigational Device Exemption (IDE) protocol for PMEG (IDE#: G200159) at our center.

Since the first EVAR in 1991, decades of collective experience have led to the next phase in its evolution in the form of hostile neck devices and branch incorporation techniques such as FEVAR. Along with the technological refinement and ever-expanding device choices, physicians have a greater responsibility to gain knowledge of failure modes, the respective bailout options, and most importantly better understanding of appropriate patient selection. Therefore, acquisition of knowledge and skills required in fenestrated endovascular rescue can provide a useful armamentarium for treating physicians.In conclusion, fenestrated endovascular repair is feasible for proximal failure of Ovation endografts. Careful planning and advanced skill set in complex endovascular aortic repair are required, as well as detailed knowledge of the failed endografts. Adjunctive deployment of Palmaz stents to remediate type 1a endoleaks after the Ovation device should be carefully considered with the possible need for future revision in mind.