Thoracic Endovascular Aneurysm Sealing of a Complex Thoracoabdominal Aneurysm

Introduction

Thoracic endovascular aortic repair (TEVAR) emerged in the 1990s, reducing surgical risk and widening therapeutic indications to include high surgical risk patients, ¹ with proven short-term advantages compared with traditional open repair. ² However, the durability of TEVAR is limited by technique-specific complications during follow-up, which often require secondary endovascular interventions. ³ Standard endovascular solutions are not universally applicable to these often complex aortic morphologies, and custom-made or fenestrated endovascular aneurysm repairs (f-EVAR) are of limited utility in urgent or emergency cases due to long preparation and manufacturing times.^4,5

Recently, some authors have proposed alternative techniques that preserve visceral and renal blood flow,^6–10 including endovascular aneurysm sealing (EVAS) with the Nellix endosystem (Endologix, Irvine, CA, USA) for juxtarenal or short-neck infrarenal aneurysms. ¹¹ EVAS is currently approved for abdominal aortic aneurysm treatment and to our knowledge only 1 case of chimney EVAS for a (type V) thoracoabdominal aneurysm has been reported. ¹²

The case reported here illustrates successful thoracic EVAS (TEVAS) treatment of a type IV thoracoabdominal aortic aneurysm formed from the proximal migration of a failed fenestrated stent-graft repair; the plan was to use a single Nellix stent in the descending thoracic aorta and parallel stents to revascularize the superior mesenteric artery (SMA) and both renal arteries.

Case Report

A 74-year-old man was diagnosed with a type IV thoracoabdominal aneurysm (TAAA) at the 2-year surveillance imaging following f-EVAR of a 60-mm juxtarenal aneurysm with a 32-mm fenestrated Anaconda stent-graft (Vascutek, Renfrewshire, Scotland) and Advanta V12 stent-grafts (Atrium Medical Corp/Maquet Getinge Group, Hudson, NH, USA) for bridging the SMA and renal arteries. The previous computed tomography angiography (CTA) performed at the 1-year follow-up had confirmed aneurysm exclusion and sac shrinkage (maximum 42-mm aneurysm diameter; Figure 1A). At 2 years, the >10-mm downward migration of the proximal part of the Anaconda graft caused a significant type Ia endoleak; the Advanta stents in the SMA and renal arteries were patent and intact (Figure 1, B and C). A large focal aortic dilatation was seen proximal to the top ring of the Anaconda graft, which was floating in the 45-mm aorta (Figure 1C). Patient comorbidities included diabetes, severe chronic obstructive pulmonary disease (COPD), mild renal insufficiency (serum creatinine 1.4 mg/dL), previous venous thromboembolism (treated with warfarin), multiple coronary artery stents, and partial lung resection for cancer. The anesthetist classified the patient unfit for open conversion.

OPEN IN VIEWER

Figure 1.

(A) Computed tomography angiography (CTA) at 1 month after the successful implantation of an Anaconda fenestrated endograft with bridging stents for the superior mesenteric artery and both renal arteries; complete juxtarenal aneurysm exclusion with patent stents was obtained. (B, C) CTA 2 years later showed >10-mm downward proximal endograft migration, causing a significant type Ia endoleak, with patent and intact stents. A focal 45-mm aortic dilatation is evident at the level of the proximal 32-mm fenestrated endograft main body.

All standard endovascular solutions and custom-made devices were rejected because the protruding stents of the previously deployed Anaconda device precluded an appropriate seal zone. Planning considered the use of off-the-shelf devices including deployment of a Nellix stent extending from an undiseased proximal portion of the descending thoracic aorta into the stented part of the Anaconda stent-graft at the level of the renal arteries, possibly covering the stenotic ostium of the celiac trunk. Two covered stents were to be placed at the proximal sealing zone according to the chimney technique and extended to the right renal and SMA arteries, essentially telescoping the previous Advanta stent-grafts. The left renal artery was to receive a bare stent in the periscope configuration extending 1 cm above the Anaconda neobifurcation (Figure 2, A and B).

OPEN IN VIEWER

Figure 2.

(A) The Nellix stent is deployed from the first available normal tissue in the descending thoracic aorta, extending into the stented area of the Anaconda endograft (at the level of the renal arteries), with intentional coverage of the celiac trunk. Covered chimney stents extend from the descending thoracic aorta to the right renal and superior mesenteric arteries, “telescoping” the previous Advanta stents. A bare stent extends from the abdominal aorta up to the left renal artery (periscope configuration). (B) Distal placement of the Nellix stent 1 cm above the Anaconda neobifurcation and 12 mm below the left renal artery. (C) Postdilation using 6-×150-mm balloons for the Viabahn stents and a 5-×60-mm balloon for the S.M.A.R.T. stent. The balloons were inflated during polymer filling of the Nellix stent. (D) The presence of the proximal ring of the Anaconda graft created a kink in the stent deployed for the superior mesenteric artery. *Reinforcement with a 9-×80-mm S.M.A.R.T. stent was necessary.

The procedure was performed in a hybrid room equipped with an Artis Zee ceiling-mounted angiography system (Siemens AG, Berlin, Germany). The patient was treated under general anesthesia, and a 10,000-unit dose of intravenous heparin was administered intravenously. After surgical exposure of both the axillary and femoral arteries, a 4-F sheath was placed at the left axillary artery for a pigtail angiography catheter, which was used to monitor the surgical steps. This ready-to-use adjunctive access was also in case any rescue maneuvers were required. Next, a 70-cm, 8-F Flexor Ansel Guiding Sheath (Cook Medical, Bloomington, IN, USA) was introduced through the right axillary artery and advanced into the right renal artery on a 0.035-inch Rosen guidewire (Cook Medical) to support the advancement of a 7-×150-mm Viabahn stent-graft (W.L. Gore & Associates, Flagstaff, AZ, USA) to the target vessel. A second 70-cm, 8-F Flexor sheath was inserted into the SMA on a 0.035-inch Rosen guidewire from a second right axillary artery access to deliver an 8-×150-mm Viabahn stent-graft to the target vessel. Selective intraoperative angiography demonstrated optimal collateralization from the SMA, and given the severe stenosis at the celiac trunk ostium, a decision to intentionally cover the celiac trunk with the endobag was taken. From the right femoral access, a 45-cm, 6-F multipurpose Destination peripheral guiding sheath (Terumo, Tokyo, Japan) was inserted into the left renal artery on a 0.035-inch Rosen guidewire to support the advancement of a 7-×60-mm S.M.A.R.T. stent (Cordis Corporation, a Cardinal Health company, Milpitas, CA, USA) to the target vessel. All 3 vessels were cannulated by telescoping the previous Advanta stent-grafts of the fenestrated Anaconda device.

From the left groin, the proximal tip of a single 140-cm-long Nellix stent was advanced for proximal sealing just above the dilated thoracic aorta, approximately at the level of the ninth dorsal vertebra. The Nellix distal stent was placed 1 cm above the neobifurcation of the Anaconda and ~12 mm below the left renal artery (Figure 2B). The Viabahn stent-grafts and the S.M.A.R.T. stent were deployed first; postdilation was performed with 6-×150-mm balloons and a 5-×60-mm balloon, respectively. The Nellix stent was then dilated at 7 atm; with all 4 balloons inflated, the Nellix endobag was filled with 100 mL of polymer (Figure 2C) injected through the inline pressure transducer (up to 180 mm Hg). Prefilling was not performed because of the increased risk of endobag perforation or damage from the multiple stents in the aorta and visceral vessels, along with the need to reduce visceral occlusion time.

The time required for completing the procedure until complete curing of the polymer was 12 minutes (visceral occlusion time). Angiography confirmed complete aneurysm exclusion, absence of any endoleaks, and patency of all stents. A kink was seen in the Viabahn anchored in the SMA due to the presence of the Anaconda ring, so the Viabahn was reinforced with a 9-×80-mm S.M.A.R.T. stent (Figures 2D and 3A). A final completion Dyna-CT scan assured the absence of any significant abnormalities (Figure 3, B and C). Total operative time was 3 hours and 40 minutes, fluoroscopy time was 33 minutes, and 110 mL of iodine contrast was used.

OPEN IN VIEWER

Figure 3.

(A) Angiography confirmed complete aneurysm exclusion, absence of any endoleaks, patency of all stents, and a kink in the Viabahn stent for the superior mesenteric artery caused by the proximal ring of the Anaconda graft. (B) Final completion dynamic computed tomography showed patency of all visceral vessels and renal arteries at the axial projection. (C) No endoleak was evident at the level of the celiac trunk. (D) Computed tomography angiography (CTA) at 6 months confirmed aneurysm exclusion, endograft and stent patency without kinking, and the absence of endoleak. (E) The CTA scan illustrates the adhesion of the polymer-filled endobag to the outer curve of the visceral chimney stent and sealing of the 2 aortic focal aortic dilatations compared with preoperative imaging (F).

The patient was transferred to the intensive care unit for perioperative monitoring (3 days). The immediate onset of paraparesis was corrected with cerebrospinal fluid drainage (20 mL) combined with hemodynamic stabilization and hemotransfusion; the patient was able to walk autonomously and was discharged at postoperative day 13 with a serum creatinine of 1.2 mg/dL. Right leg paresthesia persisted for 2 months. At 6 months, a CTA scan confirmed appropriate exclusion of the aneurysm, stent patency without kinking, and the absence of endoleaks (Figure 3D-F). Over 12 months, 3 hospital readmissions were recorded associated with preexisting COPD and congestive heart failure; an echo-Doppler scan at 1 year showed patent SMA and renal arteries and a completely excluded aneurysm (a CTA was not performed at this time due to patient comorbidities). The patient is currently on lifelong acetylsalicylic acid (100 mg/d) and warfarin, with an international normalized ratio between 2 and 3.

Discussion

Complex aortic pathologies are increasingly approached with chimney, snorkel, or sandwich techniques^5–10 in association with endovascular aneurysm repair (ch-EVAR). Often these techniques are employed to obtain adequate proximal and distal landing zones, using covered stents deployed in aortic side branches parallel to and outside of the main abdominal endograft. Such off-the-shelf device use can offer urgent total endovascular solutions in potentially life-threatening situations.

Current ch-EVAR literature has demonstrated promising clinical and radiological results but has been associated with the risk of “gutter” endoleak between the chimney stents and the aortic endograft at the landing zones. ¹³ The complex interactions between the compression of the endograft and the stents, the type of stents selected, and aortic endograft oversizing seem to be subtle but of importance to ongoing durability.^5–9

It has been suggested that the chimney technique with EVAS (ch-EVAS) could reduce these complications by surrounding the chimney stent(s) with the polymer-filled endobag while maintaining stable proximal neck sealing and avoiding type II endoleak. ¹¹ Unlike traditional endografts, EVAS creates a motionless periendograft environment at the aortic sealing zone and the aneurysm sac, preventing external stent compression, ¹¹ thereby potentially increasing the spectrum of complex aortic pathologies suitable for off-the-shelf endovascular treatment.

Preliminary experiences have demonstrated the short-term efficacy of EVAS and ch-EVAS for the treatment of complications associated with failed EVAR for suprarenal and juxtarenal aneurysms, ¹⁴ especially for urgent/emergent repair where custom-made endografts cannot be deployed. ⁷ The current case demonstrates the successful TEVAS treatment of a type IV thoracoabdominal aortic aneurysm following failed f-EVAR. However, this singular TEVAS experience underlines some critical issues regarding this technique.

The Nellix stent was designed for the abdominal aorta. The instructions for use recommend proximal and distal necks ≥1 cm for adequate sealing. The authors suggest that this recommendation should also be respected in any TEVAS applications. However, in this case, a distal landing zone of undiseased aorta was unavailable, so a distal landing zone was created within the fenestrated stent-graft, with a periscope stent to the left renal artery, similar to the technique already used in pararenal aneurysms. ¹¹

The Nellix device is also recommended for treatment of an aorta <28 mm in diameter. The mean thoracic aortic diameter is often larger than 28 mm, and its adaptability to this larger diameter is yet to be seen. Further, the introducer sheath of the Nellix stent has a working length of 45 cm and can therefore easily reach the mid descending thoracic aorta. The authors suggest that a further 20 cm of proximal thoracic aorta could be achieved by advancing the Nellix stent “unprotected” from the pulled back introducer sheath, according to the normal procedures for other TEVAR devices without introducer sheaths.

Generally, it is recommended that the Nellix stent is deployed in a kissing configuration using two 10-mm-diameter Nellix stents. In the current case, as deployment was not intended to distally extend into the iliac arteries, a single device was chosen with separate parallel stents for the SMA and renal arteries. This single device configuration was selected for multiple reasons. The narrowest portion of the existing f-EVAR graft was 16 mm in the intrarenal area. Space was required for the deployment of the multiple stents for the SMA and renal arteries. Combined with the maximum diameter of the thoracoabdominal aorta (45 mm), these elements together suggested that this anatomy could accommodate a single Nellix stent plus the parallel stents and the endobag. Moreover, this single configuration can be likened to an axillobifemoral bypass, which has a single inflow graft of generally 7 mm and should not provoke problems of perfusion, which was confirmed at the 6-month CTA examination in our patient.

In the current case report, long parallel stents were employed to cannulate the SMA and renal arteries to ensure ongoing blood flow. In cases using multiple and long stents, the authors recommend that patients should be treated with anticoagulation therapy for improved long-term patency. However, there is no conclusive evidence that anticoagulation is advantageous in this setting.

The celiac trunk was covered because optimal collateralization from the SMA was demonstrated. In our experience we usually cover the celiac trunk in urgent/emergent procedures, if required, to enhance the sealing zone. In the case of inadequate collateral circulation and/or ostium stenosis, a third chimney from the left axillary access could be used.

As this is the first reported case of TEVAS with parallel stents, the results should be interpreted with caution. The authors suggest that this procedure should only be considered when open, standard, fenestrated, or branched grafts cannot be offered to patients due to polycomorbidity, inappropriate morphology, or urgent situations.

Strict follow-up of the patient into the long term is of vital importance, as associated future complications are yet to be defined. Furthermore, patients treated with TEVAS should be aware of the risks of paraparesis and paraplegia, obviously elevated in this case by the length of aorta covered by the endograft. In the current case, the use of prophylactic cerebrospinal fluid drainage might have avoided the paraparesis, ¹⁵ but the risk of spinal hemorrhage was acute in this patient given the anticoagulation therapy.

A bare stent for the left renal artery was chosen instead of a covered stent because the previous Anaconda graft might reduce the risk of endoleaks or gutters. The renal bare stent was intended simply to maintain the renal artery patency and guarantee lateral flow from the stent. Furthermore, the manipulation of the aortic arch increases the risk of stroke, and stenting of the visceral and renal vessels could increase the risk of vessel ischemia and renal impairment. ² These issues should always be discussed with the patient prior to a TEVAS intervention.

Conclusion

TEVAS with parallel stents for the visceral and renal arteries seems selectively feasible in urgent cases otherwise untreatable by other surgical or endovascular options. Long-term results and larger patient numbers are required to confirm this initial experience.