Thoracic endovascular aortic repair completion following frozen elephant trunk: how it’s done and device selection

ABSTRACT

Introduction

Since its introduction in the mid-1990s the frozen elephant trunk (FET) technique has quickly evolved into an effective hybrid treatment option for patients with various thoracic aortic pathologies, acute and chronic. However, a notable incidence of and risk for distal aortic reinterventions persists after the implementation of the FET device. In this review, the authors analyze the indications and outcomes of thoracic endovascular aortic repair completion following FET.

Areas Covered

For this review we looked not only at our own data but also searched PubMed for relevant studies, comments and current recommendations of the European Association for Cardio-Thoracic Surgery (EACTS) and the European Society for Vascular Surgery (ESVS). Additionally, we outline our approach in this 2-stage-treatment plan.

Expert Opinion

The treatment of acute or chronic aortic pathologies involving the aortic arch frequently requires a 2-stage treatment approach. Sometimes, a tertiary procedure is needed to fix the entire aortic pathology. Thoracic endovascular aortic repair completion following FET needs careful planning to achieve the excellent clinical outcomes that we and numerous other aortic centers have shown. Only a dedicated aortic clinic provides the long-term continuous follow-up required to identify the few patients in need of a tertiary procedure.

KEYWORDS:

• Aorta
• frozen elephant trunk (FET)
• distal stent graft-induced new entries (dSINEs)
• stent graft
• aortic clinic

Disclaimer

As a service to authors and researchers we are providing this version of an accepted manuscript (AM). Copyediting, typesetting, and review of the resulting proofs will be undertaken on this manuscript before final publication of the Version of Record (VoR). During production and pre-press, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal relate to these versions also.

1. Introduction

The frozen elephant trunk (FET) concept has been introduced in the mid-1990s by Kato and colleagues following the emergence of endovascular technology and availability of dedicated arch prostheses. It has quickly evolved into an excellent treatment option for acute and chronic aortic pathologies (1) involving the aortic arch including aortic dissections (type A, non-A-non-B or type B), aortic arch aneurysms, and penetrating aortic ulcers (2–7). In patients with acute or chronic aortic dissections, the FET is capable of closing the most proximal entry, thereby leading to the stabilization of the true lumen (5,8). In addition, due to the capability of the stent-graft component to expand the true lumen, the FET resolves downstream malperfusion, enables positive aortic remodeling at the level of the stent-graft and beyond and also provides an ideal artificial proximal landing zone for potential secondary downstream thoracic endovascular aortic repair (TEVAR) (9,10). Thus it has quickly evolved into an effective hybrid treatment option for patients with aortic pathologies involving the aortic arch (11), with the potential of a one-step treatment solution (12). Moreover, the FET has evolved as the preferred method for aortic arch replacement because it incorporates an anticipatory strategy aiming to be able to treat the underlying aortic pathology’s potential progression in the downstream segment (4,13,14). This is why current consensus statements by the large European societies recommend the FET for any open arch replacement procedure (15).

There have been major surgical improvements accompanying the wider use of the FET, namely the reduction of (i) surgical complexity, (ii) myocardial and visceral ischaemic times and (iii) improved perioperative outcomes due to proximalization of the distal anastomosis from zone 3 to zone 2 or possibly even more proximally (16–19). In fact, some authors have reported that FET devices routinely implanted in zone 1 or 0 (20,21). This trend towards a more proximal surgical aortic arch anastomosis, as well as a shorter stent graft length, also enhances spinal cord protection (6,22,23). Generally, FET deployment beyond the transition zones 4-5 provides a safe length for additional stent-graft deployment and easier retrograde access in case of severe aortic tortuosity (15). However, care has to be taken in order to avoid extensive covering, which is reported to be associated with an increased risk for spinal cord injury (24,25). At our center, it is standard practice to exclusively use the short version (100 mm) of the FET prosthesis resulting in almost complete absence of permanent spinal cord injury following the index procedure. The graft portion diameters range from 26 mm to 30 mm, while the stent graft portion diameters range from 28 mm to 32 mm. We use a calculated surrogate diameter to size the FET prosthesis: the anticipated predissection aortic diameters at the level of the left subclavian artery, assuming a 20% increase in aortic diameter after dissection. If there is any uncertainty, we opt for undersizing. In cases of chronic aortic dissection, where the dissection membrane is more rigid than in acute dissection, oversizing should be avoided. Therefore, for chronic aortic dissections, we always use the smallest possible stent-graft diameters of the FET prosthesis, which are 24 mm, 26 mm, or at most 28 mm.

Next to the beforementioned improvements achieved by the proximalization of the distal anastomosis there is also one major shortcoming: distal stent graft-induced new entries (dSINEs). The stent graft, by covering a smaller portion of the downstream aorta, might not properly align with the natural orientation of the aorta, potentially resulting in a sharp angle between the stent graft and the downstream aorta. This situation could lead to aberrant flow patterns towards the aortic wall, thereby increasing the risk of perforations in the dissection membrane (23). Many aortic centers have already recognized the potential clinical concern caused by dSINEs following the FET technique (23,26,27).

2. Intervention after FET

Liebrich et al showed a direct association between the rate of secondary aortic reintervention in acute and chronic thoracic aortic disease after FET procedure and a distal anastomosis in zone 2 rather than in zone 3 (6). The same is true for the shorter stent graft length (100 mm) (6). Our group has already managed to quantify the risk for secondary aortic interventions to as high as 64% after 3 years. Particularly, patients with more extensive aortic dissections or larger preoperative descending aortic diameters bear a significantly higher risk of developing distal aortic failure. (28) Moreover, diameter progression was the most common reason for aortic reinterventions. This emphasizes the advantage of distal aortic entry closure in patients with aortic dissection, as it helps to reduce persistent false lumen perfusion and aortic enlargement. Importantly, the risk encompasses both anticipated secondary aortic reinterventions and those that are truly unforeseen or unplanned. It is noteworthy that 29% of all cases involving distal stent graft extensions in our patients were indeed unexpected interventions. The various causes for reinterventions likely contributed to our inability to identify specific risk factors for aortic reinterventions in our competing risk analysis. This highlights the necessity for close and continuous follow-up of all patients undergoing the FET procedure, regardless of the underlying aortic pathology. Our patients routinely undergo follow-ups at 6 months, 12 months and annually thereafter. Preoperatively, before discharge, at each follow-up visit and whenever clinically indicated, electrocardiography-gated computed tomographic angiography scans were performed, all with slice-thickness of 3mm or less.

The FET procedure typically serves as the initial phase in a divided treatment approach in three clinical scenarios, owing to its suitability as a platform for subsequent aortic interventions.

1. Younger patients naturally face an increased risk of requiring secondary aortic interventions to post any form of aortic repair due to their longer life expectancy and greater predisposition to underlying connective tissue disorders. Therefore, in younger patients where adverse remodeling of downstream aortic segments appears probable, the FET procedure may serve as a viable surgical approach.

2. Additionally, the FET device can be employed in patients to address various aortic pathologies affecting the ascending aorta, the aortic arch, and/or the proximal descending aorta, particularly in cases where additional pathologies exist in the descending aorta but the criteria for downstream thoracic aortic interventions have not yet been met.

3. In another scenario, the FET procedure might be the preferred treatment option, but the stent-graft component may lack the necessary length to manage an additional aortic pathology spanning from the proximal to mid-descending aorta. A potential clinical illustration is an aortic dissection with a secondary sizable communication in the proximal descending thoracic aorta, which lies a few centimeters distally beyond the reach of the FET stent-graft.

Classifying these conceptual, planned or anticipated secondary aortic reinterventions after the FET procedure as a “risk” appears unwarranted. In these cases, the term “risk” for reinterventions should be used cautiously. Nevertheless, the likelihood of aortic reinterventions following the FET procedure is substantial due to unforeseen aortic behavior, stent-graft induced complications such as dSINEs development, and rare occurrences like FET graft infection (6,8,22,23,27,29).

dSINEs are known to be one major cause of unintended reinterventions following the FET procedure causing false lumen reperfusion and consequently rapid aortic growth (22,23,27,30). dSINEs may occur as a consequence of endovascular manipulation as well as interaction of the stent-graft with a fragile dissection membrane. A size mismatch between a too large stent-graft and a small true lumen seems likely to increase the risk for dSINE development. (31,32). These dSINEs functionally act as new primary entry tears. A dSINE diagnosis should be considered as a treatment failure, as mortality rates for untreated dSINEs are reported to be as high as 25% (33). Note that Kreibich et al. previously quantified a risk of dSINE development following the FET procedure of up to 25% 3 years after the index procedure, and that dSINE may develop at any time after FET implantation (23,26). For this reason, oversizing of the stent-graft portion of the FET prosthesis should be limited in patients with aortic dissections. In fact, dSINE usually develop asymptomatically and are frequently diagnosed in routine follow-up computed tomography angiography scans (8). The size and length of the FET device haven been linked with the incidence of dSINEs (27). Most available FET devices have been associated with the development of dSINEs. Hiraoka et al. showed a high incidence following the implantation of the J Graft FROZENIX (34). The same is true for Thoraflex and E-vita Open grafts (17,28). Evidence in the literature also suggests that there is an inverse relationship between dSINEs and the rate of aortic remodeling distally (23).

Another significant factor leading to unintended and urgent aortic reinterventions are thrombi within the FET stent-graft, diagnosed during the early postoperative period (30,35). Endoleaks, triggering continuous aortic enlargement typically linked with the underlying aortic pathology, are furthermore another major cause for unintended aortic reinterventions (30,36,37).

While open surgery with a conventional open anastomosis to the stent-graft portion of the FET has traditionally yielded satisfactory outcomes during the in-hospital stay, it is important to recognize several inherent shortcomings associated with this approach. Firstly, open access necessitates prolonged compression of the left lung, resulting in significant morbidity for the patient. This prolonged compression can lead to complications and may prolong the recovery process. Furthermore, there is well-documented evidence highlighting significant intraoperative bleeding originating from the stent-graft portion. This intraoperative bleeding poses challenges for surgical teams and may contribute to increased surgical complexity and postoperative complications. Additionally, postoperative damage to the stent-graft is a recognized issue, often attributed to the clamping of the stent-graft portion during the procedure. This damage can compromise the integrity of the graft and may necessitate further interventions to address complications (38,39).

If open thoracoabdominal aortic replacement is required though, clamping of the FET stent-graft is feasible, and a conventional vascular graft can be sewn to the stent-graft portion. Open surgical reinterventions are associated with acceptable perioperative morbidity and mortality, and the need for reintervention does not impact long-term survival. Close follow-up of all patients undergoing the FET procedure, regardless of the underlying disease, is essential. To prevent perioperative leakage from the FET stent-graft and improve surgical exposure by moving the anastomosis site more distally, a hybrid procedure may be a reasonable approach. Future technical advancements and new designs by FET prosthesis manufacturers may be necessary to reduce perioperative risks (e.g., spontaneous bleeding) and the incidence of aortic reinterventions (e.g., dSINE formations).

Addressing these shortcomings requires a comprehensive understanding of the underlying mechanisms and a concerted effort to develop strategies aimed at minimizing morbidity, optimizing surgical techniques, and improving patient outcomes following open surgery in conjunction with the FET procedure.

3. TEVAR

The radiopaque markers on the stent-graft component of the Thoraflex Hybrid prosthesis serve as an excellent landing platform for a distal extension in the endovascular reintervention approach.

We attempt downstream TEVAR for at least 14 days following a FET implantation in urgent scenarios. Preferably our TEVAR second stage approach is performed after full convalescence of the patients usually after 6 months. Our main intention is to maximize collateral network development in order to reduce the risk of spinal cord injury. Another precaution we take to mitigate the risk of spinal cord injury – one of our foremost concerns when addressing the descending aorta – is the routine implantation of a cerebrospinal fluid drainage the day before the TEVAR procedure, although a clear consensus regarding the prophylactic use of cerebrospinal drains is still missing (14). Note that we do not routinely implant cerebrospinal fluid drainage when implanting a FET.

TEVAR extensions are typically performed in our hybrid operating room by a member of our aortic surgical team, following current European consensus statements (15). As part of our comprehensive training in both open surgical and endovascular techniques, we prioritize the development of endovascular skills for aortic surgeons, without compromising patient outcomes. In our practice, the femoral artery serves as the primary access point for deploying the TEVAR stent-graft. We opt for a percutaneous approach, facilitating the utilization of pre-closure techniques, which streamlines the procedure and minimizes the risk of complications associated with open access methods. Once access is secured, the TEVAR stent is carefully extended down to the level of the thoraco-abdominal transition, positioning it near the origins of the coeliac trunk. This strategic placement aims to mitigate the risk of type Ib endoleaks, a recognized complication of TEVAR procedures that can compromise outcomes. To ensure optimal graft deployment and long-term efficacy, we compare the diameter of the most proximal stent graft to that of the Thoraflex hybrid prosthesis. We then implement a standard practice of oversizing the stent graft by 2 mm, providing a margin of safety and enhancing procedural success. Furthermore, we assess the diameter of the true lumen at the distal landing zone and tailor the size of the distal stent-graft accordingly, taking into account the specific characteristics of the underlying aortic pathology. For patients with classical aneurysm formation, we aim for a 10% oversizing strategy to accommodate potential changes in vessel morphology over time. Conversely, in cases of chronic dissections, we opt for tapered stent grafts and adjust the distal end sizing based on institutional standards to minimize the risk of unnecessary oversizing and associated complications. Recognizing the complexity of certain aortic pathologies, we often utilize two stent grafts to address these cases comprehensively. Among the available stent graft options, the Relay NBS Plus stent graft from Terumo Aortic, Inchinnan, UK, emerges as our preferred choice for distal stent-graft extensions in our center. This preference is rooted in its demonstrated performance and reliability in clinical settings, underscoring its suitability for managing challenging aortic conditions effectively, particularly in patients with underlying pathology of remaining dissection after previous type A repair (40). The use of the Relay NBS stent graft promotes aortic remodeling and has a low incidence of migration and minimal bird-beak and endoleak type IA (41). In the majority of cases, we use the whole 10 cm length of the FET stent-graft as a proximal landing zone. On the one hand wo want to avoid any endoleak type 1a and, on the other hand, we want to address any FET stent-graft damage that may have occurred during the index procedure (i.e. damage of the graft material by the surgical needle during the aortic anastomosis).

After the procedure we observe the patients for at least 24 hours on our intensive care unit. Here we perform continuous cerebrospinal pressure as well as invasive blood monitoring. If necessary – in case the patient develops any neurological symptoms – adjustments to the cerebrospinal pressure can be made. Next to these adjustments to the cerebrospinal pressure we also implant various other spinal cord protection measures, including elevating arterial blood pressure, ensuring adequate hemoglobin levels, utilizing fast-track concepts, and conducting serial postoperative neurological examinations. Typically, the drains are removed the following day, and patients are transferred to a standard ward for further care.

The current literature demonstrates exceptional in-hospital outcomes following downstream endovascular repair following the frozen elephant trunk procedure, characterized by notably low incidences of permanent spinal cord injury, stroke, or mortality, as depicted in Table 1.

Table 1 Overview of current literature of downstream endovascular repair following the frozen elephant trunk procedure.

Author, year	N	In-hospital mortality	Permanent SCI	Stroke
Kreibich et al., 2021 (13)	66	0 (0%)	0 (0%)	0 (0%)
Loschi et al., 2021 (26)	20	1 (5%)	1 (5%)	0 (0%)
Meisenbacher et al., 2021 (30)	20	1 (5%)	1 (5%)	2 (10%)
Haensig et al., 2020 (28)	10	1 (10%)	0 (0%)	0 (0%)
Pan et al., 2017 (29)	23	0 (0%)	0 (0%)	0 (0%)
Laranjeira Santos et al., 2017 (27)	8	1 (12,5%)	0 (0%)	0 (0%)

Yet another reason why we routinely use the Relay Plus stent graft for distal TEVAR extensions nowadays is the fact that it has a more flexible distal end and a thicker Dacron coverage compared to the rigid ring at the distal end of the Thoraflex FET stent graft portion. This makes the surgical anastomosis to the Relay Plus stent graft much easier. This is important if a 3-step procedure, namely open thoracoabdominal aortic replacement, is necessary. This 3-step approach is also made easier by the more distally located anastomosis (8).

3.1 Connective tissue disease

The use of TEVAR in patients with connective tissue diseases remains controversial. Diameter progression of the distal and proximal native aortic segments, along with complications such as membrane ruptures, has been frequently reported (42,43). An exception occurs when the proximal and distal landing zones are within a previously implanted prosthesis (43). While this primarily concerns the proximal landing zone in our experience, close follow-up enables us to detect and address potential complications early. In the European Registry of Endovascular Aortic Repair Complications (EuREC) we have not observed a case of dSINE causing malperfusion (44).

4. Conclusion

There is a significant incidence of and a risk for distal aortic reinterventions following FET implantation, both anticipated secondary aortic reinterventions and those that are truly unforeseen or unplanned, making close follow-up of patients after FET procedure independent of the underlying disease paramount, ideally in a dedicated aortic clinic. Not only does the FET prosthesis make the deployment of the TEVAR stent graft easier, it facilitates the occasionally necessary 3-step approach addressing patients carrying a high risk of negative aortic remodeling (23,30). The etiology for aortic reinterventions is multifactorial and there are no independent predictors for aortic reinterventions. Current literature has shown acceptable morbidity and mortality associated with reinterventions (30). TEVAR stent grafts are easily accessible even in emergency scenarios and attaching them on the FET prosthesis is a relatively simple procedure making this approach a relatively straightforward and safe approach to manage pathologies in the remaining descending thoracic aorta up to the level of the celiac trunk with very good postoperative outcomes and very good with no reported in-hospital mortality (45).

Another factor that has a tremendous impact on the post-operative course is the shorter operating and ventilation times and eventually (and most importantly), the left lung can remain ventilated throughout the procedure with no phases of collapse (46).

In conclusion, we and other aortic centers have shown clearly supportive data of this two-stage treatment strategy, may it be planned, anticipated, unexpected or even emergent. It addresses combined pathologies of the aortic arch and descending aorta, it minimizes surgical invasiveness during the FET implantation and maximizes spinal cord protection through the use of shorter FET stent grafts and facilitating the timely development of collateral pathways for spinal cord perfusion. In fact, we have shown no cases of postoperative permanent SCI, acute kidney failure, or in-hospital death.

5. Expert Opinion

Aortic diseases are witnessing a concerning surge in both incidence and prevalence, posing a significant challenge to global healthcare systems. The projected increase in aortic disease-related deaths by 42% until 2030 underscores the urgent need for effective management strategies. Concurrently, cardiovascular surgery societies report a steady uptick in annual aortic procedures, with a current growth rate hovering around 5% per year (47).

However, despite these alarming trends, the evidence supporting the management of aortic arch diseases remains predominantly at a ‘C’ level, presenting several challenges. One such challenge is the relatively small patient population necessitating aortic arch procedures compared to other cardiovascular conditions, albeit steadily growing. This results in low caseloads in many medical centers and limited data from published series. Moreover, the heterogeneous nature of presentations, patient profiles, and treatment approaches further complicates the development of standardized management protocols. Therapeutic decisions for aortic arch pathologies are heavily influenced by rapid technological advancements and institutional preferences, highlighting the critical need for close international scientific and clinical collaboration to address these complexities.

Identified as key areas for future clinical research, unmet needs and evidence gaps in the field of aortic arch diseases include:

• Enhancing evidence on the pathophysiology and prevention of perioperative stroke.

• Generating more robust evidence to guide the selection of optimal treatment options for patients with acute and chronic aortic arch disease.

• Furthering international standardization efforts for consistent terminology.

• Establishing standardized surveillance protocols and post-treatment follow-up strategies.

• Developing prospectively maintained, large multicentric clinical databases dedicated to aortic arch pathologies.

Despite growing interest in specialized teams to tackle aortic diseases, particularly over the past two decades, the demonstrable impact on clinical outcomes remains elusive.

Limited and controversial data characterize our understanding of secondary aortic interventions and aortic remodeling rates. For instance, Goebel et al. reported a low incidence of secondary aortic interventions following FET surgery in patients with acute dissection, possibly attributed to the limited number of computed tomographic imaging studies during follow-up (48). Conversely, Kreibich et al. demonstrated the common occurrence of secondary aortic interventions following FET surgery, irrespective of the underlying thoracic aortic pathology, with associated perioperative morbidity and mortality (30).

The introduction of the FET technique has brought to light the perioperative complication of spinal cord ischemia (SCI) in conventional aortic arch surgery. However, effective strategies to prevent SCI in conjunction with the FET procedure remain subject to ongoing debate. While landmark publications have identified specific risk factors associated with SCI, such as the distal anastomosis in arch zone 3 and certain anatomical considerations, other studies have failed to establish clear associations, indicating the involvement of alternative pathological mechanisms.

The progressive nature of underlying thoracic aortic diseases necessitates long-term considerations, despite initial successes with endografting. While endografting offers a valuable tool for addressing surgical failures in certain scenarios, long-term failures may require open surgery for definitive correction. Therefore, discussions regarding the optimal approach should consider the long-term durability of TEVAR versus surgery alongside other relevant factors.

TEVAR has significantly expanded the therapeutic armamentarium for acute and chronic thoracic aortic diseases, emerging as the preferred strategy for many cases. A deeper understanding of the natural disease course and the appropriate application of tailored strategies in specific patient cohorts have markedly contributed to improved outcomes. However, to ensure sustained progress in aortic medicine, there is an imperative for standardized surveillance protocols, comprehensive post-treatment follow-up strategies, and the establishment of large-scale clinical databases dedicated to aortic arch pathologies.

Ultimately, the establishment of aortic centers and multidisciplinary aortic teams, akin to heart teams, will play a pivotal role in advancing the field of aortic medicine to new frontiers. By encompassing the entire spectrum of treatment modalities under one collaborative umbrella, these specialized teams are poised to revolutionize the management of aortic diseases and elevate patient care to unprecedented levels of excellence.

Article Highlights

• TEVAR following FET can be a planned, anticipated, unexpected or even emergent two-stage treatment strategy.

• This two-stage treatment strategy minimizes surgical invasiveness and maximizes spinal cord protection.

• Close follow-up of patients after FET procedure, independent of the underlying disease, is paramount to identify the need for aortic reintervention

• The etiology for aortic reinterventions is multifactorial with no independent predictors.

• The FET prosthesis makes the deployment of the TEVAR stent graft easier.

• TEVAR following FET yields very good postoperative outcomes and very good with no reported in-hospital mortality.

Declaration of interest

BR performs proctor activities for Terumo Aortic. MC is a consultant for Terumo Aortic, Medtronic and Cryolife, received speaking honoraria from Bentley and is a minority shareholder of TEVAR Ltd. BR and MC are shareholders of Ascense Medical. MK has received speaking honoraria from Terumo Aortic. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewers Disclosure

Peer reviewers on this manuscript have no relevant financial relationships or otherwise to disclose.

Glossary of Abbreviations

FET	=	Frozen elephant trunk
TEVAR	=	Thoracic endovascular aortic repair
dSINEs	=	Distal stent graft-induced new entries
SCI	=	Spinal cord injury

Additional information

Funding

Funded by the Berta-Ottenstein-Programme for Advanced Clinician Scientists.