Abstract
Over the last two decades, there have been dramatic advancements in fenestrated and branched technology for endovascular repair of the thoracoabdominal aorta. The global development of these minimally invasive techniques decreased their short-term morbidity and mortality, compared to open aortic replacement. However, the need for secondary reinterventions remains frequent, although they are typically percutaneous. Continual device modifications, improved implantation techniques, and greater understanding of the endovascular tenets constantly improve rates of long-term success. Challenges exist in endovascular repair of post-dissection thoracoabdominal aortic aneurysms (PD TAAAs) with narrow paravisceral true lumen, and our institutional experience evolved from fenestrated to inner/outer branched technology to mitigate the long-term risk of target vessel instability. Refined strategies in bridging stent grafts, particularly in target vessels off the false lumen, lowered target vessel reintervention rates. Prophylactic and therapeutic management of the false lumen is essential to prevent continued aortic degeneration and to mitigate the risk of spinal cord ischemia. Over the next decade, endovascular management of PD TAAAs will continue to evolve, further improving long-term outcomes.
Keywords
Introduction
Over the last 50 years, open aortic reconstruction has been the predominant treatment approach for thoracoabdominal aortic aneurysms (TAAAs). There were significant improvements in the technique and conduct of the operation, thanks to pioneers such as Stanley Crawford 1 with subsequent advancements in left heart bypass for organ protection, 2 cerebrospinal fluid drainage, 3 intercostal reimplantation, 4 and branched surgical grafts. 5 Despite these refinements, open TAAA repair continues to be associated with significant perioperative morbidity and mortality. Even at high-volume aortic centers, contemporary literature for extent II TAAA6,7 features notable in-hospital mortality (≤9%), coupled with high rates of spinal cord ischemia (SCI; ≤14%), respiratory complications (≤42%), and renal dysfunction requiring hemodialysis (≤10%). 7
Over the past two decades, the advent and subsequent refinement of fenestrated and branched technology resulted in a dramatic increase in endovascular aortic repair (EVAR) for complex abdominal aneurysms and TAAAs, due to its reduced perioperative morbidity and mortality compared to open surgery.8–14 These endovascular techniques range from physician-modified endografts to custom-made and off-the-shelf (OTS) devices. There have been significant improvements in procedural complexity, due to design modifications in endografts11,15 and implantation techniques. 16 The robust global experience in endovascular TAAA repair led to an expeditious decrease in perioperative morbidity and mortality, with concomitant decrease in mid- to long-term aortic and target vessel (TV)-related reinterventions. As a result, current European guidelines recommend fenestrated or branched endovascular repair for suprarenal abdominal aortic aneurysms and TAAA in high-risk candidates for open surgery. 17 Herein, we discuss the tenets for endovascular success in fenestrated-branched EVAR (FBEVAR), as well as our institutional endovascular approach to post-dissection TAAA (PD TAAA).
Principles of endovascular success in TAAA
The proximal landing zone
Careful selection of proximal and distal landing zones is the first step in endovascular management of TAAA. Procedural planning requires centerline-based identification of disease-free, non-aneurysmal seal zones proximally in the aorta and distally in the infrarenal aorta (IRA) or iliac arteries. It is important to balance the proximal seal with the extent of intercostal and lumbar coverage to minimize the risk of SCI. In extent IV TAAA, a proximal landing zone at least 5 cm proximal to the celiac fenestration or branch results not only in a robust seal, but also simplifies subsequent thoracic EVAR (TEVAR) in the event of future type 1A endoleak. In extent I-III TAAA, endovascular repair should be staged, with a first-stage TEVAR and second-stage FBEVAR. Furthermore, a proximal seal is often feasible in Ishimaru zones 2–4 when the descending thoracic aorta (DTA) is <42 mm.
However, it is necessary at times to seal proximally through open, endovascular, or hybrid techniques. This clinical scenario is common after chronic PD TAAA after open ascending replacement for Stanford type A aortic dissection. Multidisciplinary discussion between vascular surgery and cardiac surgery is essential, balancing surgical risk with institutional success in redo open aortic surgery with arch repair and frozen elephant trunk and endovascular aortic arch (endoarch) repair. The former facilitates a proximal landing zone for future endovascular interventions, while the latter can seal in an ascending aortic surgical graft.
Furthermore, endoarch repair emerged as a viable alternative to open arch surgery for both degenerative and post-dissection aneurysms over the last decade.14–16 The aortic arch presents unique anatomic challenges to endovascular repair, including a durable proximal seal zone, preservation of the supra-aortic trunks through fenestrated or branched techniques, and orientation of the endograft to the outer curve. While open repair remains the gold standard for aortic arch repair, 18 many patients in today's aging population are poor surgical candidates, based on comorbidities and the risks of cardiopulmonary bypass and reoperative sternotomy. 19
The distal landing zone
The IRA or iliac system is selected for the distal seal in endovascular TAAA repair. Particularly in PD TAAA, we prefer to land in the IRA to preserve the lumbar arteries and inferior mesenteric artery (IMA). 20 This location is favorable to stage the aortic coverage and mitigate the risk of SCI. However, it is not feasible in an aneurysmal IRA or common iliac artery. In such anatomy, an EVAR with or without an iliac branch endoprosthesis (IBE) is necessary. The IBE facilitates a durable distal fixation while preserving the internal iliac artery and reducing the sequelae of pelvic ischemia.21,22
TV incorporation—fenestrations versus branches
Incorporation of renovisceral vessels is achieved using fenestrations and/or directional branches. We utilize fenestrations when there is endograft apposition to the aortic wall; others apply a more liberal approach, tolerating a gap distance between endograft fenestration and aortic wall <5 mm. 23 Long-term patency of fenestrations is particularly favorable to branches, especially for renal arteries. 24 Directional branches (Figure 1(a)–(d)) are preferred in the absence of endograft apposition to the aortic wall or presence of significant aortic or iliac tortuosity, where misalignment of the fenestrations is anticipated. Furthermore, OTS devices (i.e. multi-branch stent-graft Zenith t-Branch, Cook Medical, Bloomington, IN, USA) and physician-modified endografts are utilized in urgent and emergent repairs. The strategy of fenestrations and/or branches is individualized and carries trade-offs in terms of sealing zone length, TV incorporation, and long-term effects on TV instability. 25
(a) Three-dimensional reconstruction showing distal aneurysmal degeneration of a previously treated thoracic aortic aneurysm, now progressed to a Crawford type II thoracoabdominal aortic aneurysm (TAAA). (b) Transfemoral renovisceral stenting with three coaxial sheaths of 18F DrySeal Flex Sheath (W. L. Gore & Associates) with 33 cm length, 16F HeliFX Aptus Tourguide steerable sheath (Medtronic Inc.), and a Flexor Ansel 1 guiding sheath (8F × 90/110 cm; Cook Medical). Note the 16F steerable sheath just proximal to the branch. (c) Completion aortography with exclusion of the TAAA with the t-Branch and patency of the renovisceral branches. (d) Three-dimensional reconstruction of the postoperative computed tomography angiography at 1 month.
Furthermore, hostile renal anatomy includes small renal diameter, upward-facing vessels, and early bifurcation, and they represent anatomic challenges for endovascular preservation. First, the incorporation of renal arteries that are <4 mm is associated with worse TV-related outcomes. 26 Thus, we include renal arteries ≥4 mm in FBEVAR. Second, upward-facing renal arteries can be addressed with retrograde (upward) renal branches 27 or antegrade (downward) branches in ballerina configuration. 28 Third, early bifurcation of the renal artery can be addressed with antegrade and retrograde branches, bidirectional double inner branches, 29 or double-barrel stenting. 30 Finally, polar renal arteries ≥4 mm can be incorporated via fenestrations or branches, as previously described.
Branched endovascular repair in PD TAAAs
PD TAAAs (Figure 2(a)) represent a particularly challenging cohort for endovascular success, often featuring true lumen (TL) compression, TV arising from the false lumen (FL), and/or dissection extending into TVs. It can affect the technical success and long-term endovascular outcomes of FBEVAR. The visceral aorta often features asymmetric TL morphology with varying degrees of compression.
Three-dimensional reconstruction demonstrating (a) a patient with prior open ascending and aortic arch replacement with total cervical debranching and frozen elephant trunk, TEVAR, and infrarenal fenestration after a large angioplasty balloon. The residual type III TAAA chronic dissection of the visceral and infrarenal aorta (b) was treated with a custom-made FEVAR (Cook Medical) implanted above the IMA, and BeFlared bridging stents (Bentley, Hechinghen, Germany). The distal right polar renal artery was preserved with a 5th fenestration.
Our transition from fenestrations to outer/inner branches in PD TAAAs
We initially favored fenestrations for renovisceral incorporation to avoid extrinsic compression of the external branches by the septum. However, our practice evolved to branched EVAR (BEVAR) (predominantly OTS Cook Medical t-Branch) in hopes of mitigating TV instability of renovisceral fenestrations, particularly those associated with FL origin. In our series of BEVAR in PD TAAA, TL diameter did not affect technical success (94%) in 34 patients, and there was a median visceral TL diameter of 13 mm, 28 with 99% TV patency at a median follow-up of 18 months. TV incorporation was performed through ipsilateral femoral access (after release of the branched endograft and removal of the delivery system) in nearly two-thirds of patients. This adaptation facilitated single femoral large-bore access, with no events of spinal cord, pelvic, and extremity ischemia. Finally, we observed expansion of the branched endograft over time in the setting of a depressurized FL, despite the presence of a chronic and thickened lamella.
On the other hand, others adopted inner branch technology (OTS E-nside; Artivion, Kennesaw, GA, USA) in PD TAAA. A recent study of 34 patients demonstrated comparable technical success (97%) and TV-related outcomes (87%) at a mean follow-up of 15 months. 31 However, the preloaded inner branches required bilateral femoral large-bore access, which might contribute to increased risk of spinal cord ischemia (9%). As the European experience develops with this device, procedural refinements will likely mitigate the risk of SCI and improve TV-related outcomes.
Our strategy for bridging stents in PD TAAA
In BEVAR, we prefer balloon-expandable stent-grafts (BESG), such as BeGraft peripheral PLUS (Bentley InnoMed GmbH). However, in the setting of long (i.e. >8 cm) branches, TV off the FL, dissected TV, and/or hereditary thoracic aortic disease, we utilize self-expandable stent-grafts (SESG), such as Viabahn (W. L. Gore & Associates). It is important to always reinforce SESG to mitigate stent graft compression. A balloon-expandable bare-metal stent, such as the BeSmooth peripheral (Bentley InnoMed GmbH), can reinforce the overlap with the branch as well as the curve of the SESG all the way to the origin of the TV. This recently described28,32 hybrid stenting optimizes the benefits of a proximal BESG and a distal SESG to minimize the risk of short and long-term TV outcomes, including kinking, type Ic and IIIc endoleaks, and iatrogenic dissection of the TV.
FL management
FL perfusion results in aneurysmal degeneration of PD TAAA. 33 As a result, FL management is essential to achieve thrombosis and sac regression in PD TAAA. Furthermore, complete FL thrombosis reduces the chronic inflammatory response seen with persistent FL flow 34 and is associated with improved long-term survival. In contrast, partial FL thrombosis is linked to the highest rates of adverse aortic remodeling, with faster aortic growth. 35 The goal of the fenestrated-branched endograft is to seal all re-entry tears. However, continued FL flow can exist through type R endoleaks (i.e. intercostal and lumbar arteries, IMA, polar renal artery) or distal re-entry tears in the IRA or iliac system with retrograde flow to the thoracoabdominal FL.
To prevent the former (type R endoleaks), we liberally perform coil (Concerto Detachable Coils System, Medtronic, Minneapolis, MN, USA) or Amplatzer Vascular Plug (AVP; Abbott, St Paul, MN, USA) embolization of intercostal and lumbar arteries. 36 This adjunctive technique, also known as minimally invasive staged segmental artery coil embolisation (MISACE), 37 primes the paraspinous collateral network. The PAraplegia Prevention in Aortic Aneurysm Repair by ThoracoabdomInal Staging (PAPAartis) trial will hopefully demonstrate that this staged approach to intercostal occlusion results in decreased rates of SCI. It can be performed before or after FBEVAR. In our recent study, 28 50% of patients with PD TAAA underwent prophylactic embolization of TL or FL intercostal or lumbar arteries before (9%) and at (41%) the time of BEVAR.
To address retrograde perfusion of the thoracoabdominal aorta, we utilize FL endograft (FLE) and/or AVP or coil embolization to achieve FL thrombosis. The FL diameter determines which technique to employ. FLEs (Figure 2(b)) are ideal when the FL is <4 cm in diameter for 4 cm in length. Using centerline-based FL measurements, the largest diameter of the elliptical FL is oversized by 10%–20% to determine the appropriate diameter of the FLE. Furthermore, minimal thrombus is necessary to avoid renovisceral thromboembolization. It can be implanted in the distal DTA during staged TEVAR, or in the IRA when the distal BEVAR is proximal to the IMA. FLEs in the distal DTA and IRA result in sequential FL thrombosis to reduce SCI risk. Furthermore, it is important to align the distal stent of the FLE with the distal TEVAR or BEVAR/cuff to mitigate the risk of stent-graft-induced new entry tear. When the FL diameter is >4 cm, but with a smaller channel perfusing the FL, we utilize large AVP and coils. In our series, the majority (56%) of patients underwent FLE (38%) or AVP/coil-based FL occlusion (18%) at the time of BEVAR.
Other techniques exist to obliterate the FL, including the Knickerbocker technique 38 and thermal septotomy. 39 However, we rarely perform them in this cohort, as it can cause a conflict with TV catheterization and distal embolization.
Conclusions
FBEVAR has matured into a versatile and durable option for both chronic PD and atherosclerotic TAAAs. Constantly evolving refinements in landing zone selection, renovisceral incorporation, and perioperative strategies will further solidify its role and safety in contemporary aortic practice. Chronic PD TAAA represents a challenging cohort for endovascular success, and our institutional experience with BEVAR, particularly the OTS t-Branch, and sequential FL thrombosis through intercostal/lumbar embolization and FLE should be considered in the armamentarium of endovascular aortic interventionalists.
Footnotes
Declaration of conflicting interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: RGA is a consultant for Cook Medical. SH has intellectual property and is a consultant for Cook Medical, Bentley, and GE Healthcare.
Ethical approval and informed consent statements
This review article does not contain any studies with human or animal participants.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
