Redo Fenestrated-Branched Endovascular Aortic Repair (F-BEVAR) for Failed F-BEVAR

Introduction

Fenestrated-branched endovascular aortic repair (F-BEVAR) is becoming the first treatment option for failed infrarenal EVAR secondary to loss of proximal seal. Regarding failed F-BEVAR, the redo endovascular repair is indeed more challenging because of the endograft limbs limiting tracking and manipulation of the new device, significant metal artifact, and presence of the bridging stents in the aortic lumen.^1,2

The aim of this study was to report 2 cases of redo F-BEVAR for the treatment of type IIIb endoleaks. The devices implanted were investigational and used as part of a Physician-Sponsored Investigational Device Exemption study protocol (PS-IDE#G140108, NCT #02266719). Inclusion and exclusion criteria are available in the supplementary material (Appendix 1). Informed consent was obtained.

Methods

Case 1

A 74 year-old female with a history of thoracoabdominal aortic aneurysm underwent F-BEVAR at another institution 5 years prior to presentation. Past medical history of cardiac arrythmia, hypertension, hyperlipidemia, chronic obstructive pulmonary disease, and chronic kidney disease (GFR 33 mL/min/1.73 m² and 1.6 mg/dL creatinine). The original patient-specific device had been designed with 2 branches for the superior mesenteric artery (SMA) and celiac artery and 2 fenestrations for the renal arteries (Figure 1A). During her 4 year follow-up visit, the computed tomographic angiogram (CTA) demonstrated a posterior endoleak at the SMA level, which was confirmed to be a type IIIb (fabric tear) with dynamic volumetric CTA (DV-CTA). The new patient-specific device was designed similar to the previous one with 2 branches for the SMA and celiac artery and 2 fenestrations for the renal arteries (Figure 1B). The patient-specific device is based on a modular system constructed of full-thickness woven polyester fabric sewn to self-expanding stainless-steel z-stents with braided polyester and monofilament polypropylene sutures. The modules are fully stented to provide stability and expansible force during development, and attachment and seal of the graft to the vessel wall after deployment. All the 4 components were preloaded for access from above using the transabdominal preloaded guidewire delivery system (TPDS). The branches were located more proximal to the previous ones and the fenestrations were positioned at the same level and clock positions but within a narrower segment of the device (Figure 1B).

OPEN IN VIEWER

Figure 1.

The original patient-specific device designed with 2 branches for the superior mesenteric artery (SMA) and celiac artery and 2 fenestrations for the renal arteries (A). The new patient-specific device was designed with 2 directional branches for the SMA and celiac artery and 2 fenestrations for the renal arteries. All the 4 branches components were preloaded for access from above, using the transabdominal preloaded guidewire delivery system (TPDS) (B). IVD, internal vessel diameter.

Case 2

A 72 year-old female had undergone EVAR of an infrarenal aneurysm 4 years before. Past medical history of coronary artery disease with prior myocardial infarction and hypertension, with preserved renal function (GFR 62 mL/min/1.73 m² and 0.8 mg/dL creatinine). The patient presented with proximal degeneration and enlargement of the thoracoabdominal aortic segment with a diameter of 56 mm. Staged thoracic and subsequent with patient-specific BEVAR device with 4 branches and an inverted limb bifurcated graft (Figure 2A) were performed. On the 30-day follow up, CTA revealed an endoleak of indeterminate origin. Follow-up imaging, 3 months later, including CTA, DV-CTA and selective angiograms of the target arteries demonstrated a type IIIb endoleak from a tear in the graft posterior to the renal artery branch. A decision to perform redo BEVAR was made and a new patient-specific device with 4 branches was designed. All 4 branches were located proximal to the previous ones and preloaded for access from above using the TPDS (Figure 2B).

OPEN IN VIEWER

Figure 2.

The original patient-specific-device designed with 4 branches for the target vessels (A). The new patient-specific device designed with 4 directional branches. All the 4 branches components were preloaded for access from above, using the TPDS (B).

Results

In case 1, the total operating time was 495 minutes, fluoroscopy time was 141 minutes, and contrast volume was 90 mL. No intraoperative complications occurred. Completion angiogram showed excellent technical result with perfusion of all visceral arteries and no endoleak (Figure 3C). The postoperative course was unremarkable, and the patient was discharged home on postoperative day 3. Follow-up at 12 months showed patent device stent-graft and no endoleak (Figure 3D).

OPEN IN VIEWER

Figure 3.

Case 1. Four-year postoperative CTA 3-dimensional reconstruction demonstrated a possible type IIIa posterior endoleak at the SMA level (A). Intraoperative angiography demonstrated a type IIIb posterior endoleak at the SMA level (B). Completion angiogram showed excellent technical result with perfusion of all visceral arteries and no endoleak (C). One-year postoperative CTA 3-dimensional reconstruction demonstrated widely patent device stent-graft and no endoleak with a stable aneurysm sac (D).

In case 2, the total operative time was 405 minutes, fluoroscopy time was 58 minutes, and contrast volume was 100 mL. No intraoperative complications occurred. Completion angiogram also showed excellent technical result, with no endoleak (Figure 4C). The postoperative course was unremarkable, and the patient was discharged home on postoperative day 3. Follow-up at 6 months showed patent stent grafts and no endoleak (Figure 4D). Of importance, the fabric tear and source of the type IIIb endoleak was inadvertently catheterized while attempting to cannulate the renal artery branch, which confirmed the nature of the endoleak (Figure 5).

OPEN IN VIEWER

Figure 4.

Case 2. Three month postoperative CTA 3-dimensional reconstruction demonstrated an endoleak of indeterminate origin (A). Intraoperative angiography demonstrated a type IIIb from a tear in the graft posterior to the renal artery branch (B). Completion angiogram showed excellent technical result with perfusion of all visceral arteries and no endoleak (C). Six-month postoperative CTA 3-dimensional reconstruction demonstrated widely patent device stent graft and no endoleak with a stable aneurysm sac (D).

OPEN IN VIEWER

Figure 5.

Intraoperative fluoroscopic image demonstrating the catheter through the fabric tear (black arrow), which was inadvertently catheterized while attempting to cannulate the renal artery branch (A). Digital subtraction angiography confirming the type IIIb endoleak (black arrow) from the fabric tear on the graft (B).

Discussion

These 2 cases demonstrate that even though redo F-BEVAR is a highly complex procedure, it is a feasible treatment option for failed F-BEVAR and can be performed with satisfactory technical success and safety. Failed F-BEVAR, secondary to a type IIIb endoleak from a breakage in the main device integrity, requiring a redo F-BEVAR is a rare event, with 2% (5/221) reporting rates in prior series.^1,2 Regarding the PS-IDE study at our site, this complication represents less than 0.4%, corresponding to 1 among 247 patients that underwent F-BEVAR with at least 30 day follow-up. As mentioned above, the first case was initially enrolled at another institution. Failed F-BEVAR secondary to a type IIIb endoleak from a brakeage in the main device integrity is an extremely rare adverse event after F-BEVAR using custom made devices. ⁴ To the best of our knowledge, this is the first report in the literature of this complication. In both cases, the location of the endoleak was adjacent to the visceral vessels, therefore, no other type of intervention would have been feasible in without compromising target vessel patency. We did not consider an alternative treatment options like parallel stenting inside the F-BEVAR because we believe that the gutters present in this technique could possibly not treat the endoleak properly. Open conversion was not appropriate as both patients were unfit for open repair and explanting branched/fenestrated endografts is a formidable task with prohibitive morbidity and mortality.

Prior studies have described the feasibility of redo F-BEVAR after failure of previous FEVAR secondary to type Ia endoleak or device migration.^2,5,61 Most of these failed fenestrated EVARs have actually been treated with redesigned fenestrated devices. To our knowledge, this is the first series where redo branched devices have been used for this purpose. Larger series have reported good outcomes of F-BEVAR procedures after failed infrarenal endovascular repair. ⁷ Comparable to those cases, the redo F-BEVAR is technically challenging because of the manipulation and rotation of the F-BEVAR device, which can be limited by the previous graft. Even though femoral access with the use of steerable sheaths is gaining popularity, redo F-BEVAR may be difficult given the limited space within the constraints of previously placed fenestrated-branched device and the limited ability to rotate or displace the device longitudinally. To overcome such limitations, we use the SCAPED technique with progressive deployment, which maintains the constraining wires, significantly decreases the diameter of the endograft and allow increased working room that facilitates selective catheterization of the target vessels. A transfemoral approach with the use of steerable sheaths could be an alternative technique in cases with an upward orientation of the target vessels. ³ However, in these 2 cases, the narrow space in the paravisceral area, severe aortic angulation in Case 1 and downward orientation of the directional branches favored the used of brachial access and sequential deployment with the use of preloaded devices. Finally, the presence of an endograft, particularly one in the thoracoabdominal space, limits severely the ability to orient the device to match the fenestrations/branches with the original components of the target arteries.

The metal artifacts and multiple radiopaque markers from the prior graft increases the complexity of the redo procedure. To minimize the risk of incorrect deployment of the graft and facilitate target vessel cannulation, we routinely use IVUS and fusion imaging. The use of IVUS during fenestrated endovascular aortic aneurysm repair (FEVAR) is of particular utility as it provides live accurate imaging of the visceral segment of the aorta for the location of target vessels, identification of thrombus or calcification, and aortic sizing. Also, IVUS allows fluoroscopic marking of the visceral vessel origins which facilitates target vessel catheterization. These advanced intraoperative imaging technologies are imperative to achieve technical success in complex cases and can decrease fluoroscopic time and contrast volume. Another strategy to minimize the technical demands of this procedure is the use of preloaded delivery systems, which eliminates the need to cannulate the device and the branch/fenestration and has been reported with high technical success rates and with a trend to reduce procedural metrics. ⁸ The TPDS is a novel delivery system designed with two, 0.018 in., wires precatheterized in each branch or fenestration, which forms 2 loops in the top of the delivery system. The device uses a 20Fr delivery system with a long 8Fr nosecone that is connected to the tip of the 20Fr cannula that allows immediate access to the directional branches or fenestrations. The device is advanced via through-and-through brachial-femoral wire and the 8Fr nosecone is exteriorized via the 12Fr. The long nosecone used in the novel upper extremity delivery systems is a stable system that prevents the guidewire from wrapping and also facilitates orientation and deployment of the device. This is particularly helpful in patients with tortuous anatomy or in those with prior failed endografts.

Both patients had an unremarkable postoperative course without any complications. The 3 day total hospital course is the standard length after F-BEVAR in our institution, particularly for most extensive repairs. This protocol allows rigorous postoperative management with intravenous hydration, optimal blood pressure levels and neurological monitorization to mitigate complications in the postoperative course. Regarding the preference to not use a prophylactic spinal drain, we do not use routinely in F-BEVAR procedures. Spinal cord protection relies primarily on maintaining a perioperative systolic blood pressure between 140 and 160 mm Hg or a mean arterial pressure >90 mm Hg, avoiding hypotension, preservation of as many collateral beds as possible, staged repairs, and early lower extremity reperfusion based on neuromonitoring. Prophylactic spinal drain placement is currently reserved for high-risk patients with aortic coverage more than 5 cm above the celiac artery and incomplete collateral network. A recent study has shown that F-BEVAR can be performed with a minimal risk of spinal cord ischemia (SCI) without the need for routine prophylactic spinal drains, however, further evaluation is needed. ⁹

As we mention earlier, failed F-BEVAR requiring redo F-BEVAR is an uncommon event but cannot be disregarded considering that the endovascular repair of thoracoabdominal and complex aortic aneurysm have been more disseminated worldwide and its long-term outcomes are still being evaluated.

Conclusion

Redo fenestrated and branched EVAR after failed F-BEVAR is a feasible technique with excellent technical success and low morbidity. However, it is a technically challenging procedure that demands advanced intraoperative imaging technologies, an experienced operator and staff to be performed safely. Finally, larger series and long-term follow-up is needed to assess the efficacy and safety of this technique.

Supplemental Material