The 3-step approach for the treatment of multisegmental thoraco-abdominal aortic pathologies

Abstract

OBJECTIVES

The goal of this study was to describe our 3-step approach to treat multisegmental thoraco-abdominal aortic disease due to aortic dissection and to present our initial clinical results.

METHODS

Nine patients with multisegmental thoraco-abdominal aortic pathology due to aortic dissection underwent our 3-step approach, which consisted of total aortic arch replacement via the frozen elephant trunk technique, thoracic endovascular aortic repair for distal extension down to the level of the thoraco-abdominal transition and, finally, open thoraco-abdominal aortic replacement for the remaining downstream aortic segments. We assessed their baseline and aortic characteristics, previous aortic procedures, intraoperative details, clinical outcomes and follow-up data.

RESULTS

The median age was 58 (42–66) years; 4 patients (44%) presented connective tissue disease. Eight patients (89%) had undergone previous aortic surgery for aortic dissection. In-hospital mortality was 0% (n = 0). None suffered symptomatic spinal cord injury or disabling stroke. During the follow-up period, 1 patient died of acute biliary septic shock 6 months after thoraco-abdominal aortic replacement.

CONCLUSIONS

The 3-step approach to treat multisegmental thoraco-abdominal aortic pathology due to aortic dissection, which involves applying both open and endovascular techniques, is associated with an excellent clinical outcome and low perioperative risk. Distal shifting of the disease process through the thoracic endovascular aortic repair extension—and thereby necessitating limited open thoraco-abdominal aortic repair—seems to be the major factor enabling these favourable results.

IRB approval

IRB approval was obtained (No. 425/15) from the institutional review board of the University of Freiburg.

INTRODUCTION

Multisegmental aortic diseases are rare and are often diagnosed in patients with a residual aortic dissection after previous type A dissection repair who are lost to follow-up. Nevertheless, progression of the diameter resulting in aneurysm formation is seen in up to 60% of patients after aortic dissection, irrespective of the initial treatment modality, especially if several aortic segments between the aortic arch and aortic bifurcation are involved [1]. Once the diameter threshold of 55 mm for aortic replacement or intervention is reached in several segments, different open surgical and endovascular treatment options are available, including combining the 2 techniques [2]. Moreover, the optimal order in which to treat the aortic pathologies surgically still needs to be determined.

The goal of this study was to describe our 3-step approach for treating multisegmental thoraco-abdominal aortic pathology due to aortic dissection and to present our initial clinical results.

PATIENTS AND METHODS

Approval was obtained (No. 425/15) from the institutional review board of the University of Freiburg. Informed consent was waived because the study was a retrospective analysis.

Patients

Nine patients underwent staged treatment for multisegmental aortic diseases involving thoraco-abdominal aortic pathology due to aortic dissection at the University Heart Center Freiburg between April 2013 and February 2020. We included all patients who had undergone (i) ascending and total aortic arch replacement via the frozen elephant trunk (FET) technique, followed by (ii) thoracic endovascular aortic repair (TEVAR) and followed by (iii) open thoraco-abdominal aortic replacement (TAAR) of the remaining aorta at our centre, in that order.

Data collection and definition of parameters

Data were collected retrospectively from our centre’s aortic database. We defined multisegmental aortic diseases entailing thoraco-abdominal aortic pathology as any aortic dilatation exceeding 40 mm and involving all aortic segments between zones 0 and 5 or at least all segments up to zone 4 and more than 2 segments beyond zone 5 [3]. Non-A non-B aortic dissection is defined according the current guidelines as aortic dissection involving the aortic arch but not the ascending aorta [4, 5]. Indication for elective surgery was an aortic diameter >55 mm in the respective aortic segment or a diameter progression of >10 mm/year. Stroke was defined according to the VARC-2 criteria [6]. Baseline and aortic characteristics, previous aortic procedures, intraoperative details, clinical outcomes and follow-up data were evaluated. We defined descending aortic segments as L1, stent graft level of the FET prosthesis; L2, thoraco-abdominal transition and L3, coeliac trunk level [7].

If each aortic segment reached a diameter >55 mm at the first presentation, we would proceed as follows: FET—TEVAR within 14 days—open repair within 3 months.

Total arch replacement via the frozen elephant trunk technique (first step)

At our centre, the FET technique is routinely performed via a complete sternotomy, and the stent graft is sized according to the diameter of the true lumen without oversizing in patients with chronic aortic dissection. The subclavian/axillary artery is now used for arterial cardiopulmonary bypass inflow. The femoral or carotid artery is used if the subclavian/axillary artery is not suitable. Concomitant valve or root procedures are routinely carried out while cooling the patient to a core body temperature of 25°C. Bilateral cerebral perfusion is routine for cerebral protection. We apply cold-blood cardioplegia or the beating-heart technique using 300 ml normothermic myocardial perfusion for myocardial protection via the cardioplegia cannula during aortic arch replacement [8]. We avoid oversizing the stent graft component of the Thoraflex prosthesis in chronic aortic disease and employ the short version of the Thoraflex hybrid graft (100 mm) (Terumo Aortic, Inchinnan, UK) exclusively. The distal anastomosis is routinely done in zone 2. We oversew the stump of the subclavian artery, insert the hybrid graft under direct vision and anastomose the prosthesis as a (i) distal anastomosis, (ii) left subclavian artery, (iii) left carotid artery, (iv) proximal anastomosis and (v) brachiocephalic trunk anastomosis [9]. Indications for total aortic arch replacement were diameter progression in all patients exceeding the recommended limit of 55 mm in the distal ascending aorta and/or within the aortic arch (extending to zone 4).

Endovascular extension (second step)

The stent graft component of the Thoraflex Hybrid prosthesis (with its radiopaque markers) is an excellent landing platform for a distal extension for a subsequent TEVAR. We apply cerebral spinal fluid drainage routinely to protect the spinal cord. Our standard is percutaneous access to the common femoral artery via a preclosure technique (Proglide, Abbott Medical, Chicago, IL, USA). The TEVAR extension usually is carried out down to the level of the thoraco-abdominal transition to obtain maximum coverage and shift the proximal anastomosis for the third step more distally. The size of the distal stent graft depends on the diameter of the true lumen measured at the distal landing zone, and the proximal stent graft routinely exceeds the FET diameter by 2 mm. We frequently use tapered stent grafts to comply with the true lumen diameters as needed and to prevent oversizing in the dissected landing zone. Indications were planned extension of the FET prosthesis in 7 (78%) patients and endoleak Ib in 1 patient (11%). Because another patient had a kink in the FET prosthesis, an uncovered stent was implanted at the level of the stent graft component of the FET prosthesis. In this patient, the uncovered stent had perforated the FET prosthesis, which ultimately required a TEVAR extension. The L3 diameter before TEVAR was 41 (38–47) mm.

Open thoraco-abdominal aortic replacement (third step)

Also in this case, cerebral spinal fluid drainage is routine for spinal cord protection. A thoraco-phrenico-lumbotomy is performed, and partial CBP is installed via the left groin. The stent graft is clamped, and the proximal anastomosis is done using a Siena prosthesis in an inverted fashion (Terumo Aortic, Inchinnan, UK). If the space suffice to clamp cranially to the coeliac trunk, we do that; if not, we carry out infrarenal or iliac clamping and establish selective antegrade normothermic blood perfusion to the visceral and renal arteries as needed. The anastomotic sequence is proximal anastomosis, distal anastomosis (afterwards, upper and lower body perfusion are connected again), right renal artery, superior mesenteric artery, coeliac trunk and finally the left renal artery. If needed, segmental arteries are reimplanted accordingly, the patient is weaned from cardiopulmonary bypass, and we close the wound in layers. We pay utmost attention to keeping the left lung ventilated and disconnect it only for brief periods to improve aortic exposure. Indications were insufficient coverage of the implanted stent graft resulting in persistent false lumen perfusion and therefore impeding aortic rupture. One patient (11%) had a distal stent graft-induced new entry (dSINE) after the second step with impending rupture of the thoraco-abdominal aorta; another patient (11%) underwent TAAR for an aneurysm of a reimplanted island of the visceral and renal arteries following open TAAR years before the first admission to our centre. Six patients needed open TAAR surgery for a diameter (progression) exceeding 55 mm distal to the previously implanted stent grafts.

Technical details on Siena-to-thoracic endovascular aortic repair anastomosis

We used the sewing collar of the Siena prosthesis to correct diameter discrepancies between the stent graft (Fig. 1A) and the aortic wall serving as a taper to the main graft carrying the 4 side branches used for the visceral and renal arteries (Video 1) [10]. In certain clinical scenarios, the diameter of the distal stent graft in the true lumen was small, whereas the total aortic diameter was extremely large. Under these specific conditions, we used the remaining rim of the resected elephant trunk component of the Siena prosthesis for end-to-end anastomosis solely to the stent graft and then employed the sewing collar for a circumferential anastomosis to the outer aortic wall. We use the term ‘double anastomosis’ for this technical refinement (Fig. 1B), which is in our opinion a substantial advantage compared to other thoracic aortic aneurysm prostheses in this special scenario.

Figure 1:

(A) Intraoperative view into the descending aorta with a stent graft in situ and the remaining rim of the resected elephant trunk component of the Siena prosthesis for end-to-end anastomosis solely to the stent graft. (B) Completed distal anastomosis using the sewing collar for a circumferential anastomosis to the outer aortic wall and the entirely replaced thoraco-abdominal aorta.

Statistical analyses

All values are expressed as number (percentage) or median (first quartile–third quartile). We used IBM SPSS Statistics 24 for Macintosh (Armonk, NY, USA) for the statistical analyses.

RESULTS

Demographics

This 3-step approach was used in 9 patients (78% male). The median age was 58 (42–66) years. Connective tissue disease was diagnosed in 4 patients (44%). Five patients suffered from progression of the diameter of the residual aortic dissection at aortic segments located more distally after previous type A repair. Two further patients were treated for chronic aortic type B dissection (1 received an uncovered stent beforehand): 1 had a chronic non-A non-B dissection and the other had an acute non-A non-B dissection after valve-sparing root replacement for an aortic root aneurysm (1 had failed, and his aortic valve had to be replaced 1 year later). Patient demographics are summarized in Table 1; previous aortic procedures are shown in Table 2.

Table 1:

Patient demographics

Demographics	n = 9
Age (years)	58 (42–66)
Male	7 (78)
Hypertension	6 (67)
Chronic obstructive pulmonary disease	0 (0)
History of stroke	1 (11)
Coronary artery disease	1 (11)
Chronic kidney impairment	1 (11)
Bicuspid aortic valve	1 (11)
Connective tissue disorder	4 (44)
Chronic aortic dissection
Residual dissection after type A repair	5 (56)
Type B dissection	2 (22)
Non-A non-B dissection	2 (22)
Dissection extension
Supra-aortic vessels	6 (67)
Thoracic descending aorta	8 (89)
Abdominal aorta	8 (89)
Iliac vessels	6 (67)

Data are presented as median (first quartile–third quartile) or as number (%).

Table 2:

History of aortic disease

Previous aortic proceduresa	n = 9
Patients	8 (89)
Time since previous intervention (years)	11 (3–18)
Indication for previous aortic intervention
Acute aortic dissection type A	6 (67)
Acute aortic dissection type B	1 (11)
Ascending aneurysm	1 (11)
Previously performed aortic interventionsa
Supracoronary ascending replacement	2 (22)
With aortic root replacement	6 (67)
With hemiarch replacement	1 (11)
Previous distal aortic interventionsa	3 (33)
Uncovered descending stent	1 (11)
Open descending aortic replacement	1 (11)
Open abdominal aortic replacement	1 (11)

Data are presented as median (first quartile–third quartile) or as number (%).

^aMultiple nomination possible.

Aortic characteristics and indications for surgery

The aortic dissection affected the descending aorta in all patients and the abdominal aorta in 8 patients (89%). Figure 2 shows a preoperative computed tomography angiographic scan of a patient with multisegmental aortic disease. Aortic diameters measured from the scan after the first clinical presentation and immediately before total arch replacement are shown in Fig. 3.

Figure 2:

Preoperative computed tomography angiographic scan of a patient with a multisegmental residual aortic dissection after a previous type A repair.

Figure 3:

Aortic diameter at the stent graft level (L1), at the thoraco-abdominal transition (L2) and at the coeliac trunk level (L3) after admission and just before the first treatment step. FET: frozen elephant trunk.

Surgical characteristics

Aortic root replacement was performed in 2 (22%) patients during cooling concomitantly with total aortic arch replacement via the FET technique. The beating-heart technique for myocardial protection was used in 2 (22%) patients. For a distal TEVAR extension (second step), 1 stent graft was used in 4 (44%) patients and 2 stent grafts were used in 5 patients (56%). Surgical characteristics and intraoperative data on total arch replacement, TEVAR and open TAAR are shown in Table 3.

Table 3:

Surgical characteristics

Intraoperative data for the frozen elephant trunk procedure	n = 9
OP (min)	452 (394–559)
CPB (min)	185 (178–243)
CX (min)	105 (76–165)
SACP (min)	60 (48–63)
Lowest body temperature (°C)	24.0 (23.5–24.5)
Beating-heart technique	2 (22)
Cannulation
Femoral	1 (11)
Subclavian	7 (78)
Carotid	2 (22)
Concomitant procedures
Aortic root conduit	2 (22)
TEVAR
Time after FET procedure (days)	274 (192–665)
Number of stent grafts
1	4 (44)
2	5 (56)
OP (min)	70 (60–126)
Open TAA
Time after TEVAR (days)	281 (119–943)
OP, min	397 (357–545)
CPB, min	155 (103–205)
CX, time	120 (84–148)
Temperature (°C)	35.9 (35.0–36.6)
Distal anastomosis
Proximal to coeliac trunk	1 (11)
Oblique infrarenal	2 (22)
Infrarenal	3 (33)
Previous Y-prosthesis	1 (11)
Iliac arteries	2 (22)

Data are presented as median (first quartile–third quartile) or as number (%).

CPB: cardiopulmonary bypass; CX, cross-clamp; FET: frozen elephant trunk; OP: operation; SACP, selective antegrade cerebral perfusion; TAA: thoracic aortic aneurysm; TEVAR: thoracic endovascular aortic repair.

Clinical outcomes

The 30-day mortality was 0% (n = 0). We have not observed any symptomatic spinal cord injuries or any disabling strokes to date. One patient suffers from deranged fine motor skills that have not resolved. Re-exploration for bleeding was necessary in 1 patient (11%) after total arch replacement and in 1 patient (11%) after open TAAR. During the follow-up period, 1 patient died of acute biliary septic shock 6 months after TAAR. The median follow-up was 206 (132–745) days after completion of the third step. Clinical outcome and follow-up data are summarized in Table 4.

Table 4:

Clinical outcome and follow-up data

Clinical outcome	FET	TEVAR	TAAR
30-Day deaths	0 (0)	0 (0)	0 (0)
Disabling stroke	0 (0)	0 (0)	0 (0)
Non-disabling stroke	1 (11)	0 (0)	0 (0)
Symptomatic SCI	0 (0)	0 (0)	0 (0)
Bleeding	1 (11)	0 (0)	1 (11)
Dialysis	0 (0)	0 (0)	0 (0)
Visceral malperfusion	0 (0)	0 (0)	1 (11)
Tracheotomy	0 (0)	0 (0)	1 (11)
Follow-up data
Deaths			1 (11)
Visceral malperfusion			1 (11)

Data are presented as number (%).

FET: frozen elephant trunk; SCI spinal cord injury; TAAR: thoraco-abdominal aortic replacement; TEVAR: thoracic endovascular aortic repair.

DISCUSSION

The 3-step approach to treating multisegmental thoraco-abdominal aortic pathology due to aortic dissection using both open and endovascular techniques is associated with excellent clinical outcomes and low perioperative risk. Shifting the disease process distally via TEVAR extension—thereby reducing the invasiveness of the open thoraco-abdominal aortic repair to a bare minimum—seems to be the major factor that has enabled these encouraging results.

The numbers of patients presenting multisegmental thoraco-abdominal aortic pathology due to aortic dissection have risen in recent years due to better survival rates after type A dissection and the increased awareness of chronic and residual aortic disease. As our study results show, most of these patients had already undergone aortic surgery and suffered from the progression of the diameter of a residual aortic dissection after type A repair in more distal aortic segments. After a median of 11 years, we diagnosed a multisegmental aortic disease involving all aortic segments in our study patients. Even if there is initially no indication for replacing all aortic segments per se, that may become necessary after each procedure or because of diameter progression of untreated segments or due to complications such as impending rupture, endoleaks or dSINE [11, 12].

Although 3 of the patients in the study had already undergone surgical procedures in the distal aortic segments [(i) an uncovered descending stent in type B aortic dissection, (ii) open TAAR and (iii) infrarenal abdominal replacement], our treatment concept for patients with multisegmental aortic disease is basically the 3-step approach from proximal to distal. The FET technique is the standard procedure for total aortic arch replacement at our centre because it offers an optimal landing zone for the distal reinterventions that are often necessary [13] and shows favourable results especially in patients with a residual aortic dissection after previous aortic repair or surgery [14, 15]. The surgical complexity of the FET procedure is comparable to that of the conventional elephant trunk (cET) procedure but provides better aortic remodelling as well as a more suitable platform for distal extension and is therefore our standard technique for aortic arch replacement. Consequently, the surgical FET technique has replaced the cET technique in many cases, a procedure we reserve for rare situations such as a narrow true lumen in the thoraco-abdominal junction or when the visceral or renal arteries depend on false-lumen perfusion without further distal communications. In these extremely rare scenarios, the dissection membrane can be resected deep into the descending aorta before the cET is implanted (which is not performed by using the FET procedure) to ensure distal blood flow, and the cET can be the more suitable option. Another therapeutic option for these patients is endovascular aortic arch repair using fenestrated or branched endoprostheses. The previously implanted root or ascending prosthesis offers an ideal landing zone, and distal extension is also possible after endovascular aortic arch repair. The latest data also suggest that patients suffering from residual aortic dissections after previous type A repair can be considered patients who experience optimal clinical outcomes after endovascular arch repair with fewer neurological complications compared to those with degenerative aortic arch diseases [16]. Nevertheless, we think some issues are unresolved, i.e. the lack of long-term evidence for this relatively young patient population, the unacceptably high stroke rates and configuration in the proximal landing zone, often rendered unsuitable in everyday clinical practice due to too-short ascending replacements or substantial kinking of the prostheses. This procedure therefore remains the exception, not the rule. Moreover, TEVAR may potentially trigger negative cardiac remodelling, resulting in worsening left and right ventricular function due to the stiffer endovascular grafts (compared to the native aortic wall) [17].

Thanks to the optimal landing zone of the FET prosthesis, distal endovascular extension using TEVAR has become the treatment of choice at our centre. It has the advantage of being rapidly available even in emergency situations and, as indicated by our data, it usually delivers an excellent clinical outcome. Previous reports have shown the feasibility of TAAR following the FET procedure with good clinical results [18]. Nevertheless, implanting 2 endoprostheses just above the coeliac trunk enables us to move the proximal anastomosis of the subsequent open TAAR more distally. Patients with chronic lung diseases benefit especially from this approach, thanks to its shorter operating and ventilation times and eventually (and most importantly), the left lung can remain ventilated throughout the procedure with no phases of collapse—a factor that has a tremendous impact on the postoperative course [19]. In addition, there are fewer short-term complications such as cardiac, renal and neurological complications (especially spinal cord injury) after TEVAR compared to open replacement, and long-term complications were virtually eliminated due to further distal TAAR [13, 20]. These results were also reflected in our clinical experience when we used to replace the thoraco-abdominal aorta immediately distal to the stent portion of the FET prosthesis without previous endovascular extension (that was before we began taking the 3-step approach). Note that there are differences in feasibility and in the ease with which a conventional Dacron prosthesis is anastomosed to a stent graft: This anastomosis works well with the Relay Plus stent graft, but less with the stent graft portion of the Thoraflex prosthesis because of the design and shape of the respective stent grafts. Whereas the Thoraflex prosthesis has a rigid distal ring, the Relay Plus stent graft has a thicker Dacron part and is more flexible and therefore more effectively avoids any leakage of the anastomosis.

There is ongoing controversy regarding the use of TEVAR in patients with connective tissue diseases. Diameter progression of the distal and proximal native aortic segments, as well as specific complications such as membrane ruptures, has been frequent [21, 22]. The one exception is when the proximal and distal landing zones are within a previously implanted prosthesis [22]. Although this situation applies primarily to the proximal landing zone in our concept, we can identify and treat potential complications at an early stage through close follow-up, regardless of the fact that, in this study, 1 patient had a covered rupture due to a dSINE at the distal end of the stent that was corrected successfully in emergency surgery. We have observed no other case of dSINE causing malperfusion either in our series or in the European Registry of Endovascular Aortic Repair Complications (EuREC) registry [12].

Completion of the third step means treatment of the entire aorta from the aortic root beyond the aortic bifurcation. Because the proximal anastomosis is so challenging, we were prompted to develop the ‘double-anastomosis’ technique—one that enables us to correct diameter discrepancies effectively. The supra-aortic vessel and the perfusion branches of the Siena prosthesis can then be used for the visceral and renal arteries [7].

Endovascular repair using fenestrated prostheses is also conceivable in patients without connective tissue disease. However, because of the long waiting times for custom-made prostheses, their unavailability in emergency scenarios, the high anatomical demands (which already preclude this treatment option in many patients) and the many patients with connective tissue diseases, we have not implemented this option so far in patients who have had the FET procedure.

We observed no symptomatic spinal cord injuries in any of our patients at any step. We routinely carry out cerebral spinal fluid drainage during both TEVAR extension and open TAAR but not during the FET procedure. The ‘spinal-cord fasciotomy’ mechanism is optimally addressed by using this approach. Priming the paraspinal collateral network and anticipating the presence of the intraspinal collateral network seem to be the major contributors to these positive results with regard to spinal cord injury [23]. Maintaining collateral inflow at all steps (preservation of the left subclavian artery, selective reimplantation of thoracic or lumbar segmental arteries, preserving hypogastrics and avoiding functional occlusion during TEVAR) is another important component to keep any residual risk for spinal cord injury to a minimum [24].

Limitations and strengths

This is a retrospective single-centre study with only a small number of patients and all the inherent limitations. Moreover, there is no control group to compare with our results. However, this study describes a clear and promising concept by which to treat the entire disease and keep collateral injury to a minimum.

CONCLUSIONS

A 3-step approach to treat multisegmental thoraco-abdominal aortic pathology due to aortic dissection via both open and endovascular techniques is associated with excellent clinical outcomes and low perioperative risk. Shifting the disease process distally via a TEVAR extension—and in so doing, minimizing the need for open thoraco-abdominal aortic repair—seems to be the major factor enabling such positive outcomes. The fact that a substantial part of the descending aorta was covered by stent grafts and not entirely replaced makes periodic follow-up essential to detect possible unfortunate events as soon as they appear.

Conflict of interest: Martin Czerny is a consultant to Terumo Aortic and Medtronic, received speaking honoraria from Bentley and Cryolife and is shareholder of TEVAR Ltd. Bartosz Rylski is a consultant to Terumo Aortic.

ABBREVIATIONS

cET

Conventional elephant trunk

dSINE

Distal stent graft-induced new entry

FET

Frozen elephant trunk

TAAR

Thoraco-abdominal aortic replacement

TEVAR

Thoracic endovascular aortic repair

Author contributions

Tim Berger: Conceptualization; Data curation; Formal analysis; Methodology; Visualization; Writing—original draft. Maximilian Kreibich: Conceptualization; Supervision; Validation; Writing—review & editing. Bartosz Rylski: Conceptualization; Methodology; Supervision; Writing—review & editing. Stoyan Kondov: Data curation; Visualization; Writing—review & editing. Albi Fagu: Data curation; Writing—review & editing. Friedhelm Beyersdorf: Conceptualization; Supervision; Writing—review & editing. Matthias Siepe: Conceptualization; Supervision; Writing—review & editing. Martin Czerny: Conceptualization; Methodology; Supervision; Writing—original draft; Writing—review & editing.

Reviewer information

Interactive CardioVascular and Thoracic Surgery thanks Aung Oo and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.