Branched stented anastomosis frozen elephant trunk repair reduces operative time and transfusion requirements in acute type A aortic dissection

Abstract

Background

In acute type A aortic dissection (AADA) patients, total arch replacement (TAR) using the frozen elephant trunk (FET) technique has been adopted in emergencies to promote distal false lumen (FL) thrombosis. However, TAR is technically demanding and is associated with prolonged operative time and increased intraoperative blood transfusion. We previously reported the branched stented anastomosis frozen elephant trunk repair (B-SAFER) technique. Comparative data versus conventional TAR (cTAR) adjusted for disease severity remain limited. Here we compared the early surgical outcomes of TAR using B-SAFER versus cTAR in AADA.

Methods

This retrospective, multicenter, nonrandomized study included patients with AADA and a patent FL who underwent emergency TAR using B-SAFER (n = 63) and cTAR (n = 75). Propensity score matching based on EuroSCORE and JapanSCORE identified 48 matched pairs. After matching, 16 patients (33%) in the comparator group underwent conventional TAR without an FET device because of institutional practice patterns or device-related constraints.

Results

The 30-day mortality was 7.3% overall (7/96), with no significant difference between the groups (6.2% for B-SAFER vs 8.3% for cTAR; P > .999). Spinal cord ischemia occurred in 0% versus 4.2% (P = .153). All operative time metrics were significantly shorter with B-SAFER (all P ≤ .024). Transfusion requirements for red blood cells, fresh frozen plasma, and platelets were all significantly lower (all P < .001).

Conclusions

In propensity score–matched patients with AADA, TAR using B-SAFER was safe and reproducible, with reduced operative times and transfusion requirements compared to cTAR.

Graphical Abstract

Graphical summary of the impact of the branched stented anastomosis frozen elephant trunk repair (B-SAFER) technique on perioperative outcomes in patients with acute type A aortic dissection compared with conventional total arch replacement (cTAR). In risk score–matched patients, B-SAFER, characterized by proximalization of the distal anastomosis, selective arch branch reconstruction using branch stents, and reduced distal arch manipulation, may be associated with shorter total operative time, lower red blood cell transfusion requirements, and more ventilator-free days at 30 days than cTAR, while maintaining comparable early mortality. No left subclavian artery stent occlusion was observed.

B-SAFER may reduce operative time and transfusion and improve early recovery.

Central Message.

Total arch replacement by B-SAFER in acute type A aortic dissection may reduce operative time and transfusion requirements compared with conventional frozen elephant trunk repair with good early outcomes.

Perspective.

Prolonged operative time and excessive transfusion mainly contribute to morbidity in emergency total arch replacement for acute type A aortic dissection. By simplifying arch branch reconstruction, B-SAFER may mitigate these risks. This approach may facilitate safer and more reproducible arch repair and expand the applicability of frozen elephant trunk–based strategies in acute aortic surgery.

Background

Emergency surgical repair remains the standard treatment for Stanford type A acute aortic dissection (AADA), but it continues to be associated with high morbidity and mortality. The primary goal is to save the patient while minimizing surgical risk.1, 2, 3 Total arch replacement (TAR) for AADA is technically demanding, and prolonged operative time often increases blood transfusion and complications, significantly impacting outcomes.⁴

With improved early survival, late complications such as distal aortic dilatation and rupture have become increasingly important. A patent false lumen (FL) in the distal aorta after initial repair is a strong predictor of distal reoperation.⁵ The frozen elephant trunk (FET) procedure was introduced to promote FL thrombosis and stabilize the distal aorta. Although it has been widely adopted, TAR with FET remains technically demanding and has been associated with higher perioperative complication rates compared to partial arch replacement, raising ongoing safety concerns.6, 7, 8, 9, 10

Recently, the site of elephant trunk anastomosis (conventional or frozen) has shifted proximally from zone 3 to zone 2, to avoid the risks of deep operative fields.11, 12, 13 Although this strategy reduces technical difficulty, complex reconstruction under hypothermic circulatory arrest (HCA) remains necessary, and long operative times and perioperative complications persist. When arch branch reconstruction is required, additional anastomoses increase the complexity of the procedure.

A key conceptual strategy in contemporary arch surgery is distal anastomotic proximalization, which allows for distal repair in a more accessible operative field while potentially reducing technical complexity. However, even with proximalization to zone 2, multiple direct sutured branch anastomoses are often required, contributing to procedural complexity and prolonged HCA exposure.

To address these limitations, the branched stented anastomosis frozen elephant trunk repair (B-SAFER) technique was originally developed and reported by Roselli and colleagues.¹⁴ At our institutions, we have adopted and implemented this strategy for the management of AADA. By enabling branch revascularization using stented grafts, this strategy simplifies distal arch reconstruction. The shortening of operative and circulatory arrest times is considered a secondary consequence of this technical simplification rather than the primary objective.

We hypothesized that B-SAFER would reduce operative time and transfusion requirements compared with conventional TAR (cTAR). In addition to evaluating early operative safety and device patency, we assessed mid-term outcomes during the available follow-up, including aortic-related death and unplanned aortic reintervention. Accordingly, this multicenter retrospective analysis was undertaken to comprehensively evaluate early surgical outcomes and mid-term durability during available follow-up of B-SAFER in patients with AADA, and to directly compare these outcomes with those of cTAR.

Methods

Patients

This retrospective multicenter cohort study used data from 3 institutions collected between 2014 and 2023. During this period, 138 patients underwent emergency repair of AADA with a patent FL using TAR. TAR was performed using either an FET-based strategy or conventional arch replacement without an FET device, depending on institutional practice patterns and device availability during the study period. B-SAFER was performed in 63 patients (46%) without the need for cervical artery exposure. The mean age was 65 ± 12 years, and 46 patients were men (73%) (Table 1).

Table 1.

Baseline characteristics before and after risk score matching

Variable	Total			Risk score-adjusted
	B-SAFER	cTAR	Effect size	B-SAFER	cTAR	Effect size
Effect size
Patients, n	63	75		48	48
Age, y, mean (SD)	63 (12)	60 (13)	0.248	65 (12)	59 (13)	0.435
Male sex, n (%)	46 (73)	37 (49)	0.241	32 (67)	32 (67)	0.000
Comorbidities, n (%)
Diabetes mellitus	8 (13)	3 (4)	0.160	3 (6)	2 (4)	0.047
Dyslipidemia	23 (37)	21 (28)	0.091	16 (34)	12 (25)	0.106
Coronary artery disease	1 (2)	6 (8)	−0.144	0 (0)	2 (4)	−0.145
Cerebrovascular disease	7 (11)	3 (4)	0.137	6 (13)	2 (4)	0.151
COPD	5 (8)	7 (9)	0.025	5 (10)	4 (8)	0.036
Hypertension	49 (78)	60 (80)	−0.027	35 (73)	38 (79)	−0.073
Chronic kidney disease	24 (39)	18 (24)	0.155	16 (34)	11 (24)	−0.006
Marfan syndrome	0 (0)	2 (3)	−0.111	0 (0)	2 (4)	−0.145
Smoking	34 (54)	39 (52)	0.020	25 (52)	22 (48)	0.063
Dissection state, n (%)
Rupture (preoperative shock)	3 (5)	2 (3)	0.055	3 (6)	2 (4)	0.045
Malperfusion syndrome	16 (25)	17 (23)	0.032	11 (23)	10 (21)	0.025
Tamponade	5 (8)	5 (7)	0.024	4 (8)	5 (10)	−0.036
Partial thrombosis	18 (30)	14 (19)	0.126	17 (36)	9 (19)	−0.190
Severity, mean (SD)
Euro SCORE	9.3 (11.9)	14.8 (11.6)	−0.465	8.8 (8.3)	8.7 (9.0)	0.016
Japan SCORE	12.3 (12.2)	13.9 (10.5)	−0.148	12.5 (13.1)	12.2 (12.1)	0.025

Effect size was calculated as standardized mean difference for continuous variables and Cramer’s V for categorical variables. B-SAFER, Branched stented anastomosis frozen elephant trunk repair; cTAR, conventional total arch replacement; COPD, chronic obstructive pulmonary disease; SD, standard deviation.

The patients were divided into 2 groups: the B-SAFER group (n = 63) and the cTAR group (n = 75). Propensity score matching (PSM) based on the EuroSCORE and JapanSCORE identified 48 matched patients in each group (Table 1). After matching, 16 patients (33%) in the comparator group underwent conventional TAR without an FET device. The B-SAFER technique was introduced in July 2019. cTAR procedures were still performed during the same period at participating centers, resulting in partially overlapping treatment periods with the 2 approaches.

This study was approved by the Institutional Review Board of Saiseikai Utsunomiya Hospital (IRB 2020-50, approved September 25, 2020, and IRB 2025-S114, approved August 25, 2025). Written informed consent for the use and publication of clinical data was obtained from all patients in accordance with institutional requirements.

Given the limited study size, the multivariable model was intentionally parsimonious and should be interpreted as exploratory rather than confirmatory. Data were collected from electronic medical records and computed tomography (CT) imaging studies. B-SAFER was introduced in January 2020. This multicenter study included 3 institutions. At each center, 2 board-certified cardiovascular surgeons performed the procedures. All surgeons routinely performed complex aortic surgery, including AADA repair.

Devices and Procedure

Surgical technique

According to the Cleveland Clinic group proposed classification,¹⁵ B-SAFER can be performed in 4 general configurations: single anastomosis, single stent (SASS); single anastomosis, multiple stents (SAMS); multiple anastomoses, single stent (MASS); and multiple anastomoses, multiple stents (MAMS). Here procedures using a single branch stent (SASS and MASS) were classified as standard B-SAFER, and procedures incorporating multiple branch stents (SAMS and MAMS) were categorized as extended B-SAFER. At our institution, only MASS (n = 24; 38.1%) and MAMS (n = 39; 61.9%) were used during the study period (Table 2). At our institution, standard B-SAFER has been adopted; however, extended B-SAFER is used selectively in cases involving an aberrant vertebral artery origin or the need for additional procedures (eg, acute dissection).

Table 2.

Operative and procedural characteristics before and after risk score matching

Variable	Total			Risk score-adjusted
	B-SAFER	cTAR	Effect size	B-SAFER	cTAR	Effect size
Patients, n	63	75		48	48
DeBakey			0.082			0.119
I	44 (70)	54 (72)		33 (69)	38 (79)
II	10 (16)	8 (11)		9 (19)	6 (13)
III	9 (14)	13 (17)		6 (12)	4 (8)
Additional procedures, n (%)
CABG	2 (3.1)	3 (4.0)	−0.022	1 (2.0)	2 (4.1)	−0.060
Bentall	2 (3.1)	6 (8.0)	−0.101	2 (4.1)	4 (8.3)	−0.086
AVR	3 (5.0)	0 (0.0)	0.163	1 (2.0)	0 (0.0)	0.010
B-SAFER configuration (B-SAFER group only)						P value
MASS	24 (38.1)			21 (43.8)
MAMS	39 (61.9)			27 (56.3)
Frozen elephant trunk length, cm (mean ± SD)				9.6 (1.8)	9.5 (2.0)	.820
Distal anastomotic zone of FET, n (%)				1.4 (0.5)	2.2 (0.9)	<.001
Zone 1				27 (56)	10 (20)
Zone 2				21 (44)	17 (35)
Zone 3				0 (0)	21 (44)

Effect size is expressed as risk difference, calculated as the incidence in the B-SAFER group minus that in the FET group. Effect size was calculated as standardized mean difference for continuous variables and Cramer’s V for categorical variables. B-SAFER, Branched stented anastomosis frozen elephant trunk repair; cTAR, conventional total arch replacement; CABG, coronary artery bypass grafting; AVR, aortic valve replacement; MAMS, multiple anastomoses multiple stents; SD, standard deviation.

For arch repair, the aortic arch is transected in zone 1 or 2, and one or more supra-aortic branch vessels (eg, brachiocephalic artery, left common carotid artery) are reimplanted directly into a branched surgical graft. The graft is subsequently anastomosed to the composite of arch tissue and the FET using either the MASS or MAMS configuration, as part of conventional TAR.

During the study period, surgical strategies for AADA arch repair were distributed as follows: B-SAFER, 50%; TAR with FET, 33%; and TAR without FET, 17%. Hybrid zone 2 and branch-first techniques were not used. The operative strategy evolved over time from TAR without FET to TAR with FET, and subsequently to B-SAFER. DeBakey type II dissections were treated by standard ascending repair, whereas cases involving the distal arch or proximal descending aorta were treated by TAR with FET or B-SAFER. B-SAFER was preferentially selected when anatomically appropriate and when a left subclavian artery (LSA) stent graft was available. Final procedural selection was determined by the operating surgeon based on anatomic findings, intraoperative assessment, and device availability.

The device components used during these operations consisted of a commercially available FET device (Frozenix; Japan Lifeline) deployed into the distal arch and descending thoracic aorta. After deployment, the spatial relationship between the arch branches and the stent graft was carefully assessed, and a fenestration (∼80% of the branched graft diameter) was created using electrocautery to accommodate the target vessel. The slightly undersized fenestration facilitates controlled expansion and promotes secure sealing following balloon dilation.

Branch revascularization was achieved using medium-sized covered stents (Viabahn; W. L. Gore & Associates) introduced over the wire through the fenestrations into the target vessels. The branch stent was deployed with approximately 15 mm extending into the aortic lumen to ensure adequate anchoring. The branching device was then gently dilated using an angled clamp to minimize junctional leakage, followed by balloon dilation to achieve secure apposition. Depending on individual anatomic characteristics, additional fenestrations and branch stents were placed in the left common carotid artery or an anomalous left vertebral artery, as required. The proximal end of the branched stent graft was secured to the greater curvature of the aortic arch with a 5-0 polypropylene mattress suture (Table 3).

Table 3.

Highlights of the B-SAFER (branched stented anastomosis frozen elephant trunk repair) technique

1. Avoid oversizing of the stent graft.

2. Extend the stent 15 to 20 mm into the aortic lumen to ensure adequate anchoring.

3. Exercise caution with the left vertebral artery and limit the left subclavian artery stent to approximately 2 cm.

4. Carefully manage cases in which the vertebral artery originates directly from the aortic arch.

5. Create a fenestration approximately 80% of the graft diameter using electrocautery.

6. In cases of subclavian artery tortuosity, advance the stent over a guidewire into the true lumen before deployment.

7. Crimp the subclavian artery stent through the lumen with an angled clamp to prevent leakage at the junction of the stent and frozen elephant trunk, then insert a selective cerebral perfusion catheter and secure the stent graft by balloon dilation.

8. Fix the subclavian artery stent to the greater curvature using a 5-0 monofilament suture, positioning it proximal to the lumen when the overhang is excessive.

A conventional surgical graft (Gelweave [Terumo] or J-Graft [Japan Lifeline]) was then sutured directly to the aortic stent graft and the surrounding native aortic wall in the arch, then anastomosed proximally to the aortic root. Replacement of the aortic root was performed when indicated, depending on the extent of disease and the surgeon’s preference (Table 2).

Selection of Distal Anastomotic Level (Zone 1 vs Zone 2)

The distal anastomotic level was determined by the extent of dissection and the presence of concomitant procedures. Zone 1 was selected when root surgery was required or when the dissection involved the left common carotid artery; otherwise, zone 2 was chosen, with reconstruction of the LSA as appropriate. The number of stent grafts was determined by the extent of supra-aortic branch involvement, with multiple stents used when the dissection extended into the left common carotid artery or when direct reconstruction of the vertebral artery was required. Additionally, when aneurysmal change was present at the distal FET anastomosis, the distal landing zone was adjusted to correspond with the FET diameter, which could result in modification of the initially selected zone.

Operation

Median sternotomy was performed, and arterial cannulation was achieved via the femoral artery or directly into the ascending aorta under echocardiographic guidance (n = 54; 56.3%). After establishing cardiopulmonary bypass and systemic cooling, arch reconstruction was performed under circulatory arrest, with antegrade cerebral perfusion used in all patients. Our current institutional protocol targets a bladder temperature of 25 °C for moderate hypothermia. Historically, however, cases managed with deep HCA (DHCA) used a target temperature of ≤18 °C. These DHCA cases were performed primarily during the earlier study period in the cTAR group.

The proximal aortic arch was transected at the base of the left common carotid artery (zone 1), or between the left common carotid artery and LSA (zone 2). The FET was advanced antegrade into the descending thoracic aorta under direct vision without a guidewire. Based on the surgeon’s assessment of tissue quality or proximity of the left common carotid and LSA origins, the device was positioned in zone 1 or 2. After deployment of the FET, arch branch reconstruction was performed in accordance with the assigned technique (B-SAFER or cTAR).

The surgical graft was then beveled, with the shorter edge aligned with the aortic arch lesser curvature. For distal anastomosis, after cutting the aortic wall at zone 1 or 2, the proximal end of the inserted open stent graft was secured with a circumferential horizontal mattress suture. The surgical graft was then sutured in an end-to-end anastomosis, and the proximal anastomosis to the aortic root was completed in the standard fashion, either terminated at the sinotubular junction just above the coronary ostia or accompanied by aortic root replacement (n = 2 [3%] bioprosthetic Bentall procedures).

In the cTAR group, arch branch reconstruction was performed using conventional sutured anastomoses, without stented branch reconstruction. The distal anastomosis was constructed at zone 2 or 3 according to surgeon preference and anatomical considerations. In both groups, supra-aortic branch reconstruction was completed under selective cerebral perfusion, followed by cessation of cerebral perfusion before proximal anastomosis. After completion of the proximal anastomosis, aortic cross-clamping was released and systemic reperfusion was resumed. The strategy for supra-aortic branch reconstruction and proximal repair was identical in the 2 groups.

We reconstructed the ascending aorta and cervical artery with a woven gelatin-coated vascular prosthesis. LSA reconstruction was successfully achieved in all patients. If LSA reconstruction was technically infeasible, conversion to conventional zone 2 FET with direct LSA reconstruction was predefined as the bailout strategy.

Statistical Analysis

Categorical variables were summarized as frequencies and percentages, and continuous variables were expressed as mean ± standard deviation or median (interquartile range), as appropriate. Continuous variables were compared using the Student t test or Mann-Whitney U test, and categorical variables were compared using the test or Fisher exact test. In this limited sample, baseline characteristics were considered the primary determinants of operative burden, including operative time and the amount of blood products used. Therefore, we adjusted for differences in baseline characteristics between the 2 groups using PSM based on the EuroSCORE and JapanSCORE. Propensity scores were estimated using logistic regression to predict the use of the B-SAFER technique based exclusively on the EuroSCORE and JapanSCORE. PSM was then performed using 1:1 nearest-neighbor matching without replacement, with a caliper width of 0.20 of the standard deviation of the propensity score. Severity scores were assessed using effect size measures, including standardized mean differences (SMDs), with an SMD >10% indicating residual imbalance. Postmatching comparisons were conducted using the Student t test for continuous variables and Fisher exact test for categorical variables. Time-to-event outcomes, including survival and freedom from reintervention, were estimated using the Kaplan-Meier method, and group comparisons were performed using the log-rank test. To minimize baseline differences between groups, risk score-matched analyses were performed before outcome comparisons. All statistical analyses were performed using SPSS version 25.0 (IBM). A 2-sided P value <.05 was considered to indicate a statistically significant difference. As a sensitivity analysis, we compared operative duration and the amount of blood products between patients treated with the BE-SAFER technique and those treated with cTAR with FET. The PSM method described above was applied in the same manner.

Results

Early Outcomes

Early surgical outcomes are summarized in Tables 2 and 4. The overall operative mortality was 7.3% (7 of 96 patients), including 3 deaths (6.2%) in the B-SAFER group and 4 deaths (8.3%) in the cTAR group, with no significant between-group difference (P > .999). In the B-SAFER group, deaths were due to postoperative low-output syndrome from preoperative coronary malperfusion (n = 1), multiorgan failure (n = 1), and severe stroke (n = 1). In the cTAR group, deaths were due to low-output syndrome (n = 2; including 1 patient with Marfan syndrome and intraoperative rupture), multiorgan failure (n = 1), and severe stroke (n = 1). FET length was similar in the 2 groups (9.6 ± 1.8 cm for B-SAFER vs 9.7 ± 1.9 cm for cTAR; P = .900). Paraplegia or paraparesis occurred only in the cTAR group (n = 2), and both patients received a 9-cm FET. In the FET-only matched cohort, spinal cord ischemia (SCI) remained infrequent (0 vs 2), and no formal statistical analysis was performed because of the small number of events.

Table 4.

Surgical outcomes

Variable	Risk score-adjusted
	B-SAFER	cTAR	P value
Patients (n)	48	48
Procedure time, min, mean (SD)
Estimated arch vessel reconstruction	23.7 (14.3)	59.8 (27.7)	<.001
Total operation	332.2 (79.5)	473.8 (144.3)	<.001
CPB	196.6 (44.4)	277.2 (89.9)	<.001
Aortic cross-clamp	134.5 (33.7)	187.7 (53.0)	<.001
Cerebral perfusion	77.0 (34.3)	118.7 (43.3)	<.001
Circulatory arrest	58.1 (25.8)	69.7 (30.3)	.024
Lowest bladder temperature (°C), mean (SD)	24.5 (2.2)	22.8 (3.2)	.03
Sensitivity analysis (excluding ≤18 °C), mean (SD)∗	23.3 (2.5)		.019
Blood products (units), mean (SD)
RBC	4.5 (4.7)	13.3 (9.5)	<.001
FFP	6.8 (5.0)	15.3 (10.5)	<.001
Platelets, units, mean (SD)	16.5 (6.7)	28.1 (13.9)	<.001
Ventilator-free days at 30 d, median (IQR)	27 (7)	24.5 (28)	.035
ICU-free days at 30 d, mean (SD)†	23 (8)	22 (9)	.139
Mortality, n (%)
30-d	3 (6.2)	4 (8.3)	>.999
In-hospital	4 (8.3)	6 (13.0)	.740
Morbidity, n (%)
Cerebral infarction	1 (2.0)	3 (6.2)	.307
Spinal cord injury	0 (0.0)	2 (4.2)	.153
Permanent HD	1 (2.0)	2 (4.2)	.557
Hoarseness	0 (0.0)	1 (2.0)	.307
Reoperation for bleeding	0 (0.0)	4 (8.3)	.041
Infection	5 (10.0)	8 (16.6)	.371
Branch artery-related
Complication n (%)	0 (0.0)	1 (2.0)	>.999
ABI, mean (SD)
Right	1.12 (0.16)	1.11 (0.10)	.902
Left	1.12 (0.10)	1.15 (0.12)	.524
LSA stent occlusion, n (%)†	0 (0.0)

B-SAFER, Branched stented anastomosis frozen elephant trunk repair; cTAR, conventional total arch replacement; CPB, cardiopulmonary bypass; SD, standard deviation; IQR, interquartile range; RBC, red blood cells; FFP, fresh frozen plasma; ICU, intensive care unit; ABI, ankle-brachial index; LSA, left subclavian artery.

^∗Sensitivity analysis: To address historical changes in cooling protocols, patients managed with deep hypothermic circulatory arrest (≤18.0 °C) were excluded from the cTAR group (n = 45).

^†Mean follow-up, 36 ± 21 months.

There were significantly more ventilator-free days at 30 days in the B-SAFER group compared to the cTAR group (median [IQR], 27 [7] days vs 24 [28] days; P = .035).

Postoperative dialysis was required in 1 patient (2.1%) in the B-SAFER group and in 2 patients (4.2%) in the cTAR group. Hoarseness occurred in 0% versus 2.1% (n = 1), and reoperation for bleeding occurred in 0% versus 2.1% (n = 1); these complications did not differ significantly between the 2 groups.

Regarding operative time–related parameters, the mean operation time (332 ± 79 minutes for B-SAFER vs 473 ± 144 minutes for cTAR; P < .001), cardiopulmonary bypass time (196 ± 44 minutes vs 277 ± 89 minutes; P < .001), aortic cross-clamp time (135 ± 33 minutes vs 187 ± 53 minutes; P < .001), cerebral perfusion time (77 ± 34 minutes vs 118 ± 43 minutes; P < .001), and HCA time (58 ± 25 minutes vs 69 ± 30 minutes; P = .024) were significantly shorter in the B-SAFER group compared to the cTAR group. Concomitant procedures were infrequent: AVR (n = 1), CABG (n = 1), and the Bentall procedure (n = 2) in the B-SAFER group and CABG (n = 2) and Bentall (n = 4) in the cTAR group. After risk score matching, the rate of additional procedures remained low and comparable in the 2 groups (Table 2).

During follow-up, 16 patients (16%) underwent planned second-stage thoracic endovascular aortic repair to prevent distal stent-induced new entry, including 7 of 48 (14.6%) in the B-SAFER group and in 9 of 48 (18.8%) in the cTAR group (P = .785). Follow-up contrast-enhanced CT detected no device migration, stent fracture, component separation, or stent compression (>70% luminal stenosis). An endoleak occurred in 1 patient (2%), suspected type 1C, and resolved after planned second-stage intervention with an LSA branch stent extension. The ankle–brachial index did not differ significantly between groups (right: 1.12 ± 0.16 vs 1.11 ± 0.10, P = .902; left: 1.12 ± 0.10 vs 1.15 ± 0.12, P = .524).

The mean duration of follow-up in the B-SAFER group was 36 ± 21 months. The LSA stent patency was 100% at the latest follow-up, with no device-related stenosis, occlusion, or reintervention. All patients were maintained on single antiplatelet therapy with aspirin (Table 4).

Late Outcomes and Kaplan-Meier Analysis

The median duration of follow-up was 36 months in the B-SAFER group and 54 months in the cTAR group. Kaplan-Meier analysis showed no significant between-group differences in freedom from aortic-related death (log-rank P = .294) or freedom from unplanned aortic reintervention (log-rank P = .559) during follow-up (Figures 1 and 2). A comparison of key perioperative outcomes between the B-SAFER and cTAR groups is summarized in Figure 3.

Figure 1.

Freedom from aortic-related death after total arch replacement. Kaplan-Meier curves demonstrate freedom from aortic-related death in patients undergoing branched stented anastomosis frozen elephant trunk repair (B-SAFER) and those undergoing conventional total arch replacement (cTAR). Aortic-related death was defined as death attributable to aortic causes, and non–aortic-related deaths were censored. Tick marks indicate censoring, and numbers at risk are shown below the x-axis. Group differences were assessed using the log-rank test (log-rank P = .291). Shaded areas indicate 95% confidence limits.

Figure 2.

Freedom from unplanned aortic reintervention after total arch replacement in patients treated with branched stented anastomosis frozen elephant trunk repair (B-SAFER) and those undergoing conventional total arch replacement (cTAR). Kaplan-Meier curves demonstrate freedom from unplanned aortic reintervention, defined as unplanned endovascular reintervention during follow-up, including additional thoracic endovascular aortic repair for aneurysmal enlargement and branch-related endovascular procedures. Planned second-stage thoracic endovascular aortic repair was excluded. Tick marks indicate censoring, and numbers at risk are shown below the x-axis. Group differences were assessed using the log-rank test (log-rank P = .559). Shaded areas indicate 95% confidence limits.

Figure 3.

Comparison of perioperative outcomes between branched stented anastomosis frozen elephant trunk repair (B-SAFER) and conventional total arch replacement (cTAR) in patients with acute type A aortic dissection.

Sensitivity Analysis

A sensitivity analysis excluding patients who underwent TAR without FET yielded consistent findings. In the FET-only matched cohort (32 vs 32), baseline characteristics were comparable in the 2 groups. Operative time metrics—including total operation time, cardiopulmonary bypass time, aortic cross-clamp time, cerebral perfusion time, and HCA time—remained significantly shorter in the B-SAFER group than in the conventional FET group. Transfusion requirements also were significantly lower in the B-SAFER group (Table E1).

Discussion

This risk score–matched multicenter study demonstrates that B-SAFER is a safe, reproducible, and efficient strategy for extended TAR in patients with AADA. Compared with cTAR repair, B-SAFER achieved similar early mortality and neurologic outcomes while significantly reducing operative time, cardiopulmonary bypass time, and transfusion requirements. These differences persisted after severity adjustment, suggesting that the observed benefits are attributable primarily to procedural simplification rather than patient selection. Operative duration and ischemic time have been associated with adverse clinical outcomes in complex arch surgery.¹⁶

Despite advances in FET technology, cTAR-based arch repair remains technically demanding and frequently requires multiple deep sutured anastomoses. Large international series have consistently shown that operative and ischemic durations—rather than disease acuity itself—are associated with early outcomes.⁸

In this context, B-SAFER was developed as an extension of hybrid arch repair strategies to reduce procedural complexity while preserving distal aortic stabilization. Originally conceptualized by Roselli and colleagues, B-SAFER integrates endovascular branch reconstruction into surgical arch repair, streamlining supra-aortic vessel reconstruction into fewer stent-assisted maneuvers.^9,10,15 By minimizing sutured anastomoses under HCA, this approach facilitates completion of the distal anastomosis in a more proximal zone, which is particularly advantageous in acute dissection characterized by tissue fragility and coagulopathy. This proximalization not only improves surgical exposure but also reduces the total number of distal arch anastomoses, thereby further simplifying reconstruction in a hemodynamically unstable setting. The concept of proximalization of the distal anastomosis is also shared with the fenestrated open stent (FeneOS) technique, which aims to reduce operative invasiveness by preserving supra-aortic branch perfusion while avoiding deep operative fields and prolonged circulatory arrest.^16,17

Importantly, procedural simplification with B-SAFER did not compromise outcomes during the available follow-up period. Kaplan–Meier analysis demonstrated comparable freedom from aortic-related death in the B-SAFER and cTAR groups. When follow-up was truncated at 4 years, no significant difference was observed, suggesting that reduced operative complexity may not adversely affect survival. Aortic event–free survival, including freedom from aortic-related reintervention, also was comparable between the 2 groups, further supporting the durability of this strategy within the observed follow-up period.

Building on this conceptual framework, B-SAFER further advances the principle of proximal anastomosis by incorporating endovascular branch reconstruction, allowing the use of 1 or 2 branch stent grafts to accommodate a wide range of supra-aortic branch anatomies and pathologic conditions. This flexibility enables a secure anastomosis in a proximal zone while maintaining distal aortic stabilization, which is a key technical advantage of B-SAFER. In the present cohort, this advantage was quantitatively confirmed: the distal anastomotic zone of the FET was significantly more proximal in the B-SAFER group compared to the cTAR group (mean, 1.4 ± 0.5 vs 2.2 ± 0.9; P < .001) (Table 2).

B-SAFER was associated with substantial reductions in operative time, cardiopulmonary bypass time, cerebral perfusion time, and HCA time, accompanied by markedly lower transfusion requirements for red blood cells, fresh frozen plasma, and platelets. Given that prolonged cardiopulmonary bypass and massive transfusion are independently associated with increased mortality and major postoperative complications after acute aortic dissection repair,^18,19 these findings may reflect reduced procedural invasiveness rather than a direct survival benefit.

Neurologic safety remains a critical concern in FET-based arch repair strategies. Previous studies and meta-analyses have reported SCI rates of 4% to 7%, with longer stent graft length and more extensive distal coverage identified as key risk factors.²⁰

In the present study, SCI was rare and occurred only in the cTAR group, despite similar stent graft lengths in the 2 groups. This observation suggests that reduced ischemic duration and optimized perfusion management may contribute to neurologic safety in B-SAFER.

A potential question is why cTAR should not remain the preferred strategy for extended arch repair in AADA. cTAR remains an established and effective option, particularly in experienced centers^7,8; however, in emergency settings with acute dissection, coagulopathy, and hemodynamic instability, multiple deep anastomoses under prolonged circulatory arrest may offset the benefits of extensive distal repair. Several studies have shown that cardiopulmonary bypass time, circulatory arrest duration, and ischemic burden—rather than the extent of aortic replacement itself—predict adverse outcomes^12,13; therefore, simplifying operative steps while preserving distal aortic stabilization is critical.

From this perspective, B-SAFER may be viewed not as a replacement for cTAR but rather as a complementary strategy suited to emergency settings and patients in whom minimizing operative duration and physiologic stress is paramount. The present study suggests that B-SAFER offers a favorable balance between surgical completeness and procedural risk, achieving similar early safety with greater efficiency and faster recovery.

The applicability of B-SAFER may be particularly relevant in Japan and other Asian countries, where acute aortic dissection is more prevalent than degenerative arch aneurysm and extended arch repair is frequently required in emergency settings. In such populations, a strategy that minimizes operative time, blood loss, and physiologic stress while preserving distal aortic control is valuable. By enabling reconstruction of complex arch pathology using 1 or 2 branch stents within a multiple anastomosis framework, B-SAFER provides a pragmatic and adaptable solution for dissection-dominant disease patterns.

The most compelling finding of this multicenter retrospective study, severity-adjusted using PSM, is that B-SAFER was associated with significantly reduced operative times for AADA compared to cTAR across all measured parameters. This efficiency gain stems directly from the technical simplification inherent to B-SAFER. Originally described by Roselli and colleagues⁹ and further developed by the present authors,²¹ B-SAFER addresses a key limitation of conventional FET: the need for multiple meticulous sutured anastomoses for arch branch reconstruction under DHCA. This complexity contributes to increased procedural burden in complex aortic surgery.²²

B-SAFER enables revascularization of multiple arch branches using stented anastomoses, consolidating reconstruction into a single distal anastomosis. Our data support that this superior efficiency is reproducible across multiple centers, providing a decisive advantage over cTAR.

Reduction in Transfusion Requirements and Improvement in Systemic Outcomes

A key clinically relevant finding of this study is the association between shortened operative duration and a marked reduction in intraoperative transfusion requirements—including red blood cells, fresh frozen plasma, and platelets (all P < .001)—in the B-SAFER group. The magnitude of the operative time reduction observed in this study likely reflects not only the reduction in the number of distal arch anastomoses but also the decrease in deep tissue dissection achieved through proximalization. In acute dissection, supra-aortic branch vessels are frequently dissected and fragile, and conventional sutured reconstruction may require extensive, meticulous, and time-consuming hemostasis from friable suture lines. By avoiding multiple deep branch anastomoses, B-SAFER may substantially reduce bleeding control time in fragile dissected tissue. This effect may be particularly relevant in Asian populations, in whom smaller vessel diameters and tissue fragility can increase the technical difficulty of distal arch reconstruction and hemostasis. Prolonged cardiopulmonary bypass and surgical trauma can exacerbate coagulopathy, increasing blood loss and the need for massive transfusion.

Prolonged operative duration and ischemic time have been associated with adverse perioperative outcomes in AADA.^19,23 Additionally, massive transfusion has been associated with major postoperative complications, including acute kidney injury and acute lung injury.²⁴ Therefore, the lower transfusion requirements with B-SAFER may reflect, at least in part, reduced procedural invasiveness rather than any direct survival benefit. Patients treated using B-SAFER had significantly more ventilator-free days at 30 days and a nonsignificant trend toward more intensive care unit–free days compared to those undergoing cTAR repair.

Neurologic Safety, Strategy, and Future Directions

The early mortality rate in the B-SAFER group (6.2%) was comparable to that in the cTAR group (8.3%) and favorable relative to pooled mortality rates reported in large FET meta-analyses.^8,25,26 Meta-analytic evidence indicates that SCI risk after FET is strongly length-dependent, with devices providing coverage ≥15 cm or extending beyond T8 associated with a higher SCI rate (11.6% vs 2.5%; P < .001).⁸ This approach, combined with reduced HCA time, may increase the safety margin for spinal cord protection and contribute to the favorable neurologic outcomes observed in our cohort. B-SAFER is technically reproducible in acute settings and is associated with reduced operative burden and lower transfusion requirements.

Study Limitations

Our study has limitations inherent to a retrospective, nonrandomized, multicenter design. We adjusted for baseline characteristics between the 2 groups using risk score matching; however, other confounding factors, including anatomic variables, were not balanced, which might have affected outcomes. Procedural details and perioperative management may have varied among centers. Longer-term follow-up is needed to assess B-SAFER durability, including distal aortic remodeling and late reintervention. In addition, although the comparator group was defined as cTAR repair, 16 patients in this group underwent TAR without FET, which might have introduced procedural heterogeneity. However, a sensitivity analysis revealed that outcomes related to operative duration and blood product use were more favorable in the B-SAFER group than in the conventional TAR with FET group. Furthermore, the distal anastomotic level differed significantly between groups, reflecting a technique-specific feature of the B-SAFER procedure. Although this difference is inherent to the operative strategy, it might have influenced both operative complexity and perioperative outcomes and should be considered when interpreting the comparative results. Nevertheless, the consistent findings across centers support the robustness of the results.

Conclusions

In risk score–matched patients undergoing TAR for AADA, B-SAFER was safe, reproducible, and associated with improved procedural efficiency compared with cTAR. B-SAFER was associated with significantly shorter operative, cardiopulmonary bypass, and HCA times, resulting in a marked reduction in intraoperative transfusion requirements for all blood products. These findings support B-SAFER as a valuable strategy for mitigating major procedural risks and facilitating faster postoperative recovery in patients requiring extended arch repair.

Webcast

You can watch a Webcast of this AATS meeting presentation by going to: https://www.aats.org/resources/multicenter-severity-matched-c-12182.

Declaration of Generative AI and AI-Assisted Technologies in the Writing Process

The authors used ChatGPT (OpenAI) only for English translation and minor language refinement. The authors reviewed and edited the content as needed and take full responsibility for the scientific accuracy and integrity of the manuscript.

Conflict of Interest Statement

The authors reported no conflicts of interest.

The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.

Appendix E1

Table E1.

Surgical outcomes

Variable	Risk score–adjusted
	B-SAFER	cFET	P value
Patients, n	32	32
Procedure time, min, mean (SD)
Estimated arch vessel reconstruction	25.1 (17.1)	66.9 (26.2)	<.001
Total operation	370.1 (96.5)	488.1 (129.0)	<001
CPB	216.7 (51.0)	267.5 (59.7)	<.001
Aortic cross-clamp	148.4 (42.8)	191.4 (48.5)	<.001
Cerebral perfusion	76.5 (34.3)	128.3 (45.7)	<.001
Circulatory arrest	52.7 (14.7)	72.6 (23.3)	<.001
Blood products (utilization), mean (SD)
RBC	3.4 (4.1)	14.3 (9.2)	<.001
FFP	7.8 (6.1)	16.5 (11.3)	<.001
Platelets, units	16.4 (5.1)	26.2 (11.2)	<.001

B-SAFER, Branched stented anastomosis frozen elephant trunk repair; cFET, conventional frozen elephant trunk; SD, standard deviation; CPB, cardiopulmonary bypass; RBC, red blood cells; FFP, fresh frozen plasma.