One-year outcomes of total arch replacement and frozen elephant trunk using the E-vita Open NEO

Abstract

Open in new tabDownload slide

OBJECTIVES

In this cohort study, we aimed to assess the 1-year clinical outcomes of using the E-vita Open NEO™ hybrid prosthesis for total arch replacement with frozen elephant trunk (FET) to repair extensive aortic pathologies.

METHODS

We reviewed individuals who underwent thoracic aortic surgery between April 2021 and March 2023 from the Gangnam Severance Aortic Registry. Exclusion criteria included ascending aortic replacement, 1 or 2 partial arch replacement, descending aortic replacement and total arch replacement without an FET. Finally, all consecutive patients who underwent total arch replacement and FET with E-vita Open NEO for aortic arch pathologies between April 2021 and March 2023 were included in this cohort study. The patients were divided into 3 groups based on their pathology: acute aortic dissection, chronic aortic dissection and thoracic aortic aneurysm. The primary end point was in-hospital mortality. The secondary end points during the postoperative period comprised stroke, spinal cord injury and redo sternotomy for bleeding. Additionally, the secondary end points during the follow-up period included the 1-year survival rate, 1-year freedom from all aortic procedures and 1-year freedom from unplanned aortic interventions.

RESULTS

The study included 167 patients in total: 92 patients (55.1%) with acute aortic dissection, 20 patients (12.0%) with chronic aortic dissection and 55 patients (32.9%) with thoracic aortic aneurysm. The in-hospital mortality was 1.8% (n = 3). Strokes occurred in 1.8% (n = 3) of the patients, spinal cord injury in 1.8% (n = 3) and redo sternotomy for bleeding was performed in 3.0% (n = 5). There were no significant differences between the pathological groups. The median follow-up period (quartile 1–quartile 3) was 198 (37–373) days, with 1-year survival rates of 95.9%. At 1 year, the freedom from all aortic procedures and unplanned aortic interventions were 90.3% and 92.0%, respectively.

CONCLUSIONS

The 1-year clinical outcomes of total arch replacement with FET using the E-vita Open NEO were favourable. Long-term follow-up is required to evaluate the durability of the FET.

INTRODUCTION

The frozen elephant trunk (FET) is used in total arch replacement (TAR) to implant a stent graft into the aortic arch or descending thoracic aorta (DTA), simultaneously securing it to the vascular graft and aorta at the same time [1–7]. FET is considered the most advanced arch surgery technique because it eliminates the Ia endoleak of the stent graft while minimizing bleeding at the distal anastomosis [1–8]. It allows for the simultaneous treatment of both the aortic arch and the DTA [1–7] and facilitates potential future surgical or endovascular treatments [1–7]. Various commercial FET devices, including the E-vita™ and Thoraflex™ series, are available worldwide [3–6]. However, in South Korea, the E-vita Open NEO™ hybrid prosthesis (Artivion Inc., Kennesaw, GA, USA) is currently the only accessible FET device since April 2021.

This cohort study evaluated the outcomes of patients who underwent FET using E-vita Open NEO at a single centre over 1 year.

MATERIALS AND METHODS

We report the following article along with STROBE statement.

Ethics statement

This cohort study was approved by the Institutional Review Board (IRB No. 2023-0345-001 3-2023-0126), and the requirement for informed consent is waived.

Patients

We reviewed individuals who underwent thoracic aortic surgery between April 2021 and March 2023 from the Gangnam Severance Aortic Registry. Exclusion criteria included ascending aortic replacement, 1 or 2 partial arch replacement, descending aortic replacement and TAR without an FET. Hemi-arch replacement has not been performed in our aortic centre since 2018. Finally, a total of consecutive patients who underwent total arch replacement with frozen elephant trunk (TARFET) using the E-vita Open NEO at the Department of Cardiovascular Surgery, Gangnam Severance Hospital, Yonsei University College of Medicine between April 2021 and March 2023 were enrolled in this study (Fig. 1).

All elective surgery patients underwent preoperative pulmonary function tests, computed tomography (CT) and echocardiography, while only CT scans were performed in emergency patients.

Follow-up

After hospital discharge, patients were regularly followed up through outpatient visits. Postoperative CT scans were performed after drain removal on mean postoperative day 8 in patients with normal kidney function. Subsequent CT scans were arranged ∼6 months after discharge and then repeated every 6–12 months, depending on the patient’s condition.

Frozen elephant trunk indication

FET is typically used when there are pathologies in the ascending aorta or aortic arch along with pathologies in the DTA. The patients were classified into 3 groups: acute aortic dissection (AAD), chronic aortic dissection (CAD) and thoracic aortic aneurysm (TAA) (Fig. 2).

In the AAD group, patients with acute type I aortic dissection underwent FET when there was a tear in the distal arch or proximal DTA. As tears, whether entry or re-entry tears, can act as entry tears after surgery, we aimed to block tears located in areas coverable by FET without additional thoracic endovascular aortic repair (TEVAR). Also, FET was performed in patients with acute type I aortic dissection when there were clear symptoms such as abdominal pain or lower limb paraplegia, or if there was any portion of the distal part of the aortic arch where the true lumen had completely collapsed. Similarly, the FET was performed for acute type I intramural haematoma (IMH) when a tear or penetrating aortic ulcer in the proximal DTA was suspected. Additionally, FET is performed when the entire diameter of DTA exceeds 5 cm, either for treatment purposes or as a bridge to potential secondary procedures. In acute type III dissection/IMH cases, FET was performed for non-A non-B dissection/IMH, or when TEVAR was not feasible for complicated type B dissection/IMH.

In the TAA group, FET was used to treat an aneurysm in the ascending aorta or aortic arch as well as in the DTA in a single stage. Second-stage procedures were planned when the distal landing zone of the stent graft was positioned below T10, and FET was performed to facilitate the subsequent procedure. FET was also used in cases of a saccular aneurysm in the proximal DTA that was not amenable to TEVAR.

In the CDA group, FET was performed to tackle the dissection or aneurysm in the ascending aorta or arch while simultaneously sealing the proximal DTA under the same indication as AAD except malperfusion.

Frozen elephant trunk sizing

The FET length was primarily determined based on the outer curvature [9], aiming to land it ∼5 cm distal to the tear, PAU or aneurysm in the DTA, preferably proximal to T8. In cases where the distal landing site was below T10, a second-stage procedure was considered.

For AAD, the FET diameter matched the maximum diameter of the true lumen at the distal landing site or ∼90% of the overall maximal diameter. In the case of TAA, the FET diameter exceeded the longest diameter of the landing site by 10–20%. Conversely, for CAD, the FET diameter was increased by 10% relative to the longest diameter of the true lumen at the landing site. Different strategies have been employed based on the aortic pathology to determine the FET diameter [7].

Selection of distal anastomosis zone

One of the advantages of TARFET is the flexibility in choosing the distal anastomosis zone compared to TAR. Therefore, we first selected a device with standardized specifications for the diameter and length of the stent graft, and, based on this, determined the appropriate distal anastomosis zone. However, in the case of zone 3 anastomosis, the distal anastomosis may be somewhat more challenging than zone 1 or 2 anastomosis. Additionally, in the case of zone 1 anastomosis, the ostium of the branch grafts for head vessels can be too close to the aortic root. Therefore, we preferred zone 2 anastomosis in most cases.

Device description and device selection

The E-vita Open NEO hybrid prosthesis is a one-piece hybrid prosthesis that offers a range of vascular grafts, including straight, branched and trifurcated types, combined with a nitinol stent graft. In this device, the diameter of the vascular graft is provided in the range of 26–30 mm, the diameter of the stent graft ranges from 22 to 40 mm and the length of the stent graft is offered in the range of 120–180 mm, with slight variations depending on the type of vascular graft. Specifically, for the trifurcated type, the provided stent graft has a diameter ranging from 24 to 40 mm, with a length of 175–180 mm, offering a relatively larger and longer stent graft.

Due to the fact that all arch replacements in our centre were conducted with a branched configuration before TARFET using this device, we have a preference for the branched-type device. However, in cases requiring a large and long stent graft, there have been supply issues for branched-type devices with a large and long stent graft, domestically. In such instances, we primarily opted for trifurcated-type devices.

The diameter of the vascular graft was chosen freely in most cases, as it provides a collar length of 15 mm for the distal anastomosis.

Surgical technique

Our technique of TARFET has been described previously [7] and is similar to the TAR technique [10], while also incorporating a few distinctive strategies at our centre: (i) median sternotomy with ‘Y’ incision; (ii) arterial cannulation using the right axillary artery or right brachiocephalic artery (RBCA) with unilateral selective antegrade cerebral perfusion (SACP); (iii) moderate hypothermic circulatory arrest (MHCA) targeting at a rectal temperature of 28–30°C; (iv) no left ventricle vent cannulation; (v) ‘no-touch’ techniques for cardiac arrest; (vi) pledgeted reinforcement sutures at the proximal and distal anastomosis sites; (vii) specific order of anastomosis: distal portion of the vascular graft, left subclavian artery, left common carotid artery, the proximal portion of the vascular graft and RBCA; and (viii) Teflon felt neo-media formation technique used for patients with dissection. In our opinion, an excellent surgical view and field are the basic conditions for improving surgical outcomes. As part of this strategy, we employ unilateral SACP as well as ‘Y’ incision, no ventricle vent cannulation. However, when the regional oxygen saturation on the patient's left side decreases, indicating an incomplete circle of Willis, we additionally insert a cerebral perfusion catheter into the left carotid artery.

However, several differences exist between TARFET and TAR, which are as follows: (i) TARFET allows flexibility in placing the distal anastomosis anywhere from zone 0 to zone 3, with zone 2 being the primary location; (ii) due to this flexibility, it becomes necessary to ligate the origin site of the head vessel located more distally than the distal anastomosis for TARFET. This is achieved through the use of 4–0 Prolene, along with pledgeted 4–0 Prolene suture; (iii) TARFET involves deploying a FET device, which is typically performed with visual confirmation of its entry into the DTA. In challenging anatomies such as the tortuous aorta or the acute angle between the arch and DTA, a 035 guidewire is used for deployment and inserted via the left femoral artery puncture before the surgical incision. The orange protective flap on the delivery system was removed, and the guidewire was inserted into the over-the-wire os at the end of the shapeable shaft so that the device could be deployed along the guidewire (Fig. 3). During deployment, the first assistant held the device cuff to prevent the stent graft from rotating; (iv) to prevent cuff wrinkling, the cuff was cut ∼5 mm larger than the aorta for distal anastomosis and marked for precision.

Study definition

This study aimed to present the postoperative results and follow-up data. The cohort patients were divided into 3 study groups based on their pathology: AAD, CAD and TAA.

The primary end point was the in-hospital mortality rate. Secondary end points within the postoperative period included stroke, spinal cord ischaemia (SCI), redo sternotomy for bleeding, pneumonia, tracheostomy, bowel ischaemia, haemodialysis/continuous renal replacement therapy, sepsis and stent graft failure. Additionally, secondary end points during the follow-up period encompassed the 6-month and 1-year survival rates, 6-month and 1-year freedom from all aortic procedures and 6-month and 1-year freedom from unplanned aortic interventions, overall mortality, all aortic procedures, unplanned aortic interventions/operations and planned aortic interventions/operations.

Stroke was defined as permanent or transient hemiplegia correlated with brain lesions on CT or magnetic resonance imaging before discharge. SCI was defined as permanent or transient paraplegia or paraparesis not associated with stroke before discharge. Stent graft failure was defined as the need for additional intervention or operation due to abnormalities or problems with the stent graft portion before discharge. The term ‘procedure’ was defined to encompass both intervention and operation. Planned aortic procedures were scheduled before TARFET and were performed after discharge. In contrast, unplanned procedures occurred after discharge without prior planning. All aortic procedures involved planned and unplanned aortic procedures.

Statistical analysis

In this cohort study, the data were collected prospectively, ensuring the absence of missing data. Statistical comparisons were conducted among 3 study groups. Categorical variables were compared using the chi-squared test or Fisher’s exact test. Normality was assessed using the Kolmogorov–Smirnov and Shapiro–Wilk tests to compare continuous variables. Since the data did not follow a normal distribution, continuous variables represented as median and quartile 1–quartile 3, and the Kruskal–Wallis and Mann–Whitney post hoc tests were used for comparison of continuous variables. P-values were adjusted using Bonferroni’s method for post-hoc tests between study groups.

A risk factor analysis was performed using logistic regression for in-hospital mortality, stroke and paraplegia, respectively. Survival analysis was conducted using the Kaplan–Meier method, log-rank test and Cox proportional hazard regression for overall mortality, all aortic procedure and unplanned aortic intervention, respectively. We conducted competing risks regression considering mortality for all aortic procedures and unplanned aortic interventions among study groups. The Grambsch-Therneau test was used for assessing the satisfaction of the proportional hazards assumption. We conducted both univariable and multivariable analyses using logistic regression, Cox proportional hazard regression. A bidirectional stepwise regression approach was employed for multivariable analysis. In the univariable analysis, we included variables related to perioperative characteristics and operative details and performed multivariable analysis by including variables with P < 0.25 from the univariable analysis results. However, due to the limited number of events, caution is advised in the interpretation of multivariable analyses. Statistical significance was set at P < 0.05. Statistical analysis was performed using IBM SPSS 26.0.0.0 (IBM Corp. 2016., Armonk, NY, USA) and SAS 9.4 (SAS Institute, Cary, NC, USA).

RESULTS

All consecutive 167 patients who underwent TARFET were divided based on pathology, with 92 patients (55.1%) diagnosed as AAD, 20 (12.0%) as CDA and 55 (32.9%) as TAA. The median age of the patients was 66 (54–76) years, and there were 126 males (75.4%). When conducting a post-hoc test using Bonferroni's modification for adjusted P-values between AAD and TAA patients, statistically significant differences can be observed in baseline characteristics. Patients with AAD had undergone fewer prior cardiac or aortic procedures, along with a lower incidence of hypertension. They also had a higher proportion of emergency cases and more instances of abdominal and lower-extremity malperfusion. Conversely, patients with TAA were significantly older than those in the other groups. The baseline characteristics of the included patients are summarized in Table 1.

The basic operative strategy did not differ between the pathological groups. Axillary cannulation was performed in 103 patients (61.7%), with 2 cases using additional femoral arterial cannulation and mostly unilateral SACP (n = 156, 93.4%). The MHCA, ACC, CPB and SACP times were longer in AAD patients than in TAA patients. FET sizing varied based on pathology, resulting in differences in device size, concomitant TEVAR and extent of stent graft coverage. Patients with TAA used wider and longer stent grafts and wider vascular grafts, with a higher proportion of concomitant TEVAR and stent graft coverage extending to T8 or below. The operative details are succinctly summarized in Table 2.

Postoperative outcomes

There were no significant differences in postoperative mortality or morbidity based on the pathology (Table 3). The in-hospital mortality rate was 1.8% (n = 3), and the 30-day mortality rate was 0.6% (n = 1). Three patients died of bowel ischaemia, pneumonia and sepsis. Redo sternotomies were performed due to bleeding in 3.0% of the patients (n = 5). SCI occurred in 1.8% (n = 3); among these, 2 cases developed permanent paraplegia, and 1 case transient paraplegia. Stroke occurred in 1.8% (n = 3) of patients, with 1 case resulting in permanent hemiplegia and the other 2 resulting in transient hemiplegia. Stent graft failure occurred in 2.4% (n = 4) of patients, requiring additional TEVAR before discharge.

In terms of in-hospital mortality, the use of a trifurcated-type device was a risk factor in univariable and multivariable analysis. Furthermore, the preoperative platelet count was associated with stroke in univariable and multivariable analysis. The risk factors for SCI in univariable analysis were the intraoperative transfusion of packed red blood cells and platelet concentrates, and the intraoperative transfusion of packed red blood cells was identified as the risk factor for SCI in multivariable analysis (Supplementary Material, Table S1).

Follow-up outcomes

The follow-up results showed no significant differences between the pathological groups (Table 4 and Fig. 4). The overall mortality showed a statistically significant difference among the 3 study groups, but this difference was not significant when comparing individual study groups in post-hoc tests. The median follow-up period was 198 days (37–373 days), and the follow-up index was 0.72 ± 0.26. The 6-month and 1-year survival rates were 95.9% and 95.9%, respectively. Four deaths occurred after discharge, attributed to colon cancer, pneumonia, sepsis or an unknown cause. The freedom from all aortic procedures and unplanned aortic interventions were 97.5% and 99.2% at 6 months and 90.3% and 92.0% at 1 year, respectively. Despite employing competing risk regression, no statistically significant differences were observed based on pathology for both unplanned aortic interventions [AAD versus CAD hazard ratio = 1.43 (0.35–6.77), P-value = 0.89; AAD versus TAA hazard ratio = 1.43 (0.35–5.77), P-value = 0.62] and all aortic procedures [AAD versus CAD hazard ratio = 1.89 (0.34–10.38), P-value = 0.45; AAD vs TAA hazard ratio = 1.81 (0.49–6.71), P-value = 0.38]. During the follow-up period, a total of 11 patients required additional aortic procedure. Among them, there were no additional aortic operations; all were additional aortic interventions. Two patients planned to undergo additional aortic interventions, while the aortic interventions for 9 patients were unplanned. One of the 2 patients who underwent planned aortic intervention received a branched endovascular aneurysm repair, while the other underwent an additional TEVAR. Among the 9 patients who underwent unplanned aortic intervention, 4 had a type Ib endoleak, 3 had a retrograde endoleak, 1 had a distal stent graft-induced new entry (dSINE) and the remaining 1 was a patient with suspected right subclavian artery anastomosis leakage (Table 4). All patients with endoleaks or dSINE underwent additional TEVAR. The patient suspected of having a right subclavian anastomosis leakage did not show any leakage during angiography.

The assumption of proportional hazards was confirmed for overall mortality (P = 0.57), all aortic procedures (P = 0.90) and unplanned aortic intervention (P = 0.71) through the Grambsch-Therneau test, respectively. Coronary arterial occlusive disease (CAOD), body mass index, trifurcated-type vascular graft, stent graft diameter and stent graft length were associated with survival in univariable analysis and trifurcated-type vascular graft was identified as the risk factor for overall mortality in multivariable analysis. In univariable analysis, chronic obstructive pulmonary disease, chronic kidney disease, abdominal malperfusion and concomitant valve-sparing aortic root replacement were risk factors for both all aortic procedures and unplanned aortic interventions, and additionally, CAOD and MHCA were associated with all aortic procedures. In multivariable analysis, chronic obstructive pulmonary disease, chronic kidney disease, abdominal malperfusion and stent graft length were identified as risk factors for both all aortic procedures and unplanned aortic interventions. Additionally, in the multivariable analysis, CAOD, RBCA cannulation site and lowest rectal temperature were recognized as risk factors for all aortic procedures, and history of cardiac or aortic procedures was identified as a risk factor for unplanned aortic interventions (Supplementary Material, Table S2).

DISCUSSION

TARFET has undergone substantial evolution over time and is currently considered the most cutting-edge surgical technique in aortic arch surgery. The hybrid prostheses globally mainly used for TARFET include the E-vita series and the Thoraflex series (VASCUTEK, Terumo, Inchinnan, Scotland, UK). The E-vita Open, the first commercially available hybrid prosthesis, was launched in Europe in 2005 as a straight-type hybrid prosthesis. It evolved into the E-vita Open Plus™ in 2008, gradually enhanced with a precoated, tightening woven prosthesis, and a sewing collar. On the other hand, Thoraflex was launched in 2012 and is characterized by having a side perfusion graft and the option to use a branched-type hybrid prosthesis. In a recent meta-analysis [11], Thoraflex was reported to be favourable in terms of aortic remodelling, dSINE and endoleak. Conversely, another meta-analysis [12] suggests that E-vita performs well in 30-day mortality, in-hospital stroke and 1-year mortality. A recent original study [13] indicated that Thoraflex demonstrated advantages in postoperative outcomes, particularly in SCI, dialysis and ventilation time. In long-term outcomes, E-vita was reported to have a lower TEVAR intervention rate. In that study, the authors highlighted that the lower incidence of SCI and dialysis with Thoraflex is attributed to the presence of a side perfusion graft and the shorter length of the stent graft. Regarding the higher rate of TEVAR intervention, they pointed out the presence of a rigid ring in the stent graft. E-vita Open NEO was designed with these factors, offering a range of vascular grafts, including straight, branched and trifurcated types. Its Z-shaped nitinol stent graft offers adequate radial force while reducing distal stent-induced new entry [7, 14, 15]. Moreover, compared with other FET devices, it provides a sufficient distance from the branch graft to the sewing collar. However, there is an issue of excessive oozing from vascular grafts [16], and solutions to this challenge have been proposed and developed [7, 17].

This single-centre study aimed to evaluate TARFET outcomes in patients with extensive aortic disease. The study exhibited favourable results despite variations in baseline characteristics, indications and device size procedures based on pathology.

Compared with other studies reporting in-hospital mortality of 0–21%, stroke rates of 4–18%, SCI rates of 0–9% and redo sternotomy rates for bleeding of 3–20% for TARFET [2, 14, 18–25], our study demonstrated an overall favourable postoperative outcome.

One of the reasons for favourable in-hospital mortality may be the active use of concomitant TEVAR alongside TARFET, which has not been performed in other centres [2, 14, 18–25]. We also observed a relatively high proportion of cases with a distal landing zone at T8 or below compared to other meta-analyses [26]. In cases of TAA with extensive aneurysms, concomitant TEVAR was performed when the longest available FET was insufficient to achieve complete sealing in a single-stage operation. In most AAD cases, FET aims to prevent proximal entry or re-entry tears and address malperfusion using shorter stent grafts [2, 22]. However, in patients with AAD with a tear or aneurysm in the distal DTA, additional TEVAR was performed to exclude aortic pressure. In patients with CAD, TARFET does not aim for a single-stage operation with a longer stent graft or concomitant TEVAR. The primary objective was to exclude large entry tears from the DTA for significantly reducing the false lumen pressure. Additionally, this approach was chosen to preserve the flexibility of future options, either TEVAR or stentless TEVAR [27], if necessary. If a large entry tear was present in the distal DTA of patients with CAD, concomitant TEVAR or a longer stent graft would have been used (Fig. 2). This approach made it possible to reduce DTA rupture, as mentioned in other studies [2, 18, 21, 22] subsequently decreasing overall mortality.

Using longer stent grafts or concomitant TEVAR during TARFET can raise concerns regarding SCI. A recent meta-analysis revealed a substantial increase in SCI when the distal landing zone was at T8 or lower [26]. To mitigate this risk, whenever feasible, we positioned the stent graft above T8. However, based on our TEVAR experience, we found that cases with a distal landing at T10 or higher had a minimal occurrence of SCI. This study did not find a significant relationship between SCI and distal stent graft landing position above or below T8. Hence, in situations requiring a more distal landing, we planned the position above the T10 level to minimize the risk of SCI. For patients with stent-graft placement below T10, a second-stage procedure was planned. Three individuals in this study had SCI. Among them, 1 patient with TAA, characterized by a shaggy aorta and a distal stent graft landing zone at T8, experienced SCI due to embolization. The remaining 2 patients in the AAD group underwent concomitant TEVAR with the distal landing positions at T7 and T8. Intriguingly, all segmental arteries at the thoracic aorta level originated from the false lumen in these 2 patients, leading to rapid false lumen thrombosis and obstruction of the segmental arteries, causing SCI. This finding aligns with that of a recent study associating SCI in TARFET patients with AAD and the segmental arteries of the false lumen between T9 and L3 [28]. Thus, in AAD cases, considering the segmental artery locations when deciding on the stent graft position is crucial for reducing the risk of SCI. Of course, essentially, meticulous distal perfusion and adequate de-airing methods are vital in patients with shaggy aortas to minimize the risk of embolization leading to SCI.

We think that not only SCI but also strokes are associated with air and debris. Therefore, it is crucial to carefully de-air and eliminate any potential debris at every stage of anastomosis, both at the distal and proximal aorta as well as the head vessels, in order to mitigate postoperative neurological complications. We dedicate a substantial amount of time to this aspect.

For reducing the bleeding, it may be imperative to employ reinforcement sutures with pledgeted suture all around both distal and proximal anastomosis sites, as well as the Teflon felt neo-media formation site in the case of AAD. Additionally, it might be crucial to use 5–0 Prolene suture and plant-based haemostatic powder adequately for controlling oozing from the vascular graft.

We believe that addressing each of these major complications ultimately contributes to reducing postoperative mortality. Furthermore, we consider the aorta team members' expertise and experience in a high-volume centre to be a cornerstone in reducing mortality [25, 29]. Especially, in patients with cardiopulmonary arrest, cardiac tamponade or distal malperfusion, each team member acts promptly according to predetermined protocols, aiming to initiate surgery as quickly as possible. For instance, in emergency situations, sternotomy may commence without waiting for lines, while in parallel, our anaesthesiologists work to establish lines for infusion and arterial monitoring. We believe that this approach minimizes cardiopulmonary arrest duration, frequency and malperfusion time, contributing significantly to improved outcomes.

The follow-up data showed that the survival rate, freedom from all aortic procedures and freedom from unplanned aortic interventions were favourable compared to those in other studies [2, 18, 21–24] owing to the combination of TARFET with concomitant TEVAR, minimizing the need for future procedures. This strategy underscores a key advantage of FET over the elephant trunk method: it is a one-stage treatment without the need for future procedures.

Our overall results were comparable to those of other studies investigating AAD, CAD and TAA [2, 22]. In terms of baseline characteristics, the ADD group exhibited a lower prevalence of hypertension and surgical/procedural history, and a higher frequency of emergency statuses and cases of malperfusion. Although not statistically significant, patients with AAD showed better postoperative outcomes and fewer aortic procedures during follow-up, similar to the findings of other studies. However, disparities were found in specific outcomes, with Hellgren et al. [22] reporting the lowest in-hospital mortality in CAD and Leone et al. [2] reporting higher mortality in TAA.

We observed some discrepancies in the preoperative features upon comparing our findings with that of Tsagakis et al. [25], who investigated the E-vita Open registries. Our study included a slightly higher proportion of male patients, along with a higher number of AAD cases and fewer CAOD cases. Additionally, a history of stroke and chronic obstructive pulmonary disease also was less common. Variations in the surgical procedures were also noted. Both studies predominantly used axillary artery cannulation. However, there were differences in the cannulation alternatives. The innominate artery was used in our study, whereas the ascending aorta was used in previous studies. Moreover, while their investigation favoured bilateral SACP, ours involved unilateral SACP. Unlike E-vita Open, E-vita Open NEO is designed with both a branched and trifurcated type, facilitating separate anastomosis of head vessels. In this study, we performed separate anastomoses of the head vessels rather than island technique. Their study reported a higher in-hospital mortality rate of 15%; in contrast, it was notably lower in our study. Furthermore, compared with their research, our study demonstrated improvements in outcomes related to SCI, stroke and redo sternotomy for bleeding. In addition to our previous multicentre study in Hong Kong [7], there have been 2 other studies on TARFET using the E-vita Open NEO device. One study involving 22 patients who presented at multiple centres in Germany and Austria reported an in-hospital mortality of 9%, a stroke rate of 18.2% and no cases of SCI [19]. The other was a series study conducted in Greece, including 6 patients, with in-hospital mortality of 1 patient (17%) and no reported cases of stroke or SCI [30].

Multivariable logistic regression and Cox proportional hazard regression were approached with caution in both conduction and interpretation due to the low event count. However, we included these results in the Supplementary Material, believing that it could contribute to interpretation. In this study, only 1 variable remained statistically significant in in-hospital mortality, stroke, SCI and overall mortality after multivariable analysis. The identification of trifurcated-type devices as a risk factor for in-hospital mortality and overall mortality might seem unnatural at first glance. As explained in the ‘Device description and device selection’ section, trifurcated-type devices were primarily used when a long and large stent graft was needed. Therefore, in this study, the frequency of using trifurcated-type devices may be associated with the patients who has a larger and longer DTA, reflecting the tortuosity of the aorta. Trifurcated-type devices may have emerged as a risk factor in relation to aortic tortuosity. The finding that preoperative platelet count is a protective factor for stroke, and intraoperative packed red blood cell transfusion amount is a risk factor for SCI, suggests that the occurrence of stroke and SCI is related to the amount of bleeding before or during surgery. In the analysis of all aortic procedures and unplanned aortic interventions, a much larger number of variables emerged relative to the event count, making it likely that statistical significance may not be achieved.

Limitations

The key limitations of this study were its single-centre design and relatively short follow-up period. Previous studies using the E-vita Open NEO device, too, had small sample sizes and lacked extensive follow-up data. In the future, larger multicentre studies with longer follow-up periods are required to evaluate the durability of TARFET further using the E-vita Open NEO device.

CONCLUSION

Despite these limitations, our study demonstrated that TARFET using the E-vita Open NEO device is a feasible and versatile approach for patients with extensive aortic pathology. Particularly, in patients with different diagnoses, appropriate sizing and strategy of the FET device can lead to favourable outcomes in terms of early and midterm follow-up results, irrespective of the pathology. However, long-term studies are required to confirm the durability of this treatment.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

ACKNOWLEDGEMENTS

The authors thank Medical Illustration & Design, part of the Medical Research Support Services of Yonsei University College of Medicine, for all artistic support related to this work.

FUNDING

The authors received no financial support for the research.

Conflict of interest: none declared.

DATA AVAILABILITY

The data that support the findings of this study are available on request from the corresponding author.

Author contributions

Chong Hoon Kim: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Writing—original draft. Tae-Hoon Kim: Conceptualization; Methodology; Validation. Ha Lee: Conceptualization; Data curation. Myeong Su Kim: Conceptualization; Data curation; Validation. Woon Heo: Conceptualization; Data curation; Validation. Kyung-Jong Yoo: Conceptualization; Supervision. Bum-Koo Cho: Conceptualization; Supervision. Suk-Won Song: Conceptualization; Methodology; Supervision; Validation; Writing—original draft; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Martin Grabenwöger, Davide Pacini, Gabriele Piffaretti and the other anonymous reviewers for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• AAD

  Acute aortic dissection

 
• CAD

  Chronic aortic dissection

 
• CAOD

  Coronary arterial occlusive disease

 
• CT

  Computed tomography

 
• DTA

  Descending thoracic aorta

 
• dSINE

  Distal stent graft-induced new entry

 
• FET

  Frozen elephant trunk

 
• IMH

  Intramural haematoma

 
• MHCA

  Moderate hypothermic circulatory arrest

 
• RBCA

  Right brachiocephalic artery

 
• SCI

  Spinal cord ischaemia

 
• SACP

  Selective antegrade cerebral perfusion

 
• TAA

  Thoracic aortic aneurysm

 
• TAR

  Total arch replacement

 
• TARFET

  Total arch replacement with frozen elephant trunk

 
• TEVAR

  Thoracic endovascular aortic repair