We retrospectively analyzed 92 consecutive patients who underwent surgical treatment for ATAAD between April 2023 and March 2025. Emergency surgery was performed immediately after the diagnosis was confirmed using contrast-enhanced computed tomography (CT) of the chest and abdomen. Intraoperatively, arterial cannulation was guided by transesophageal echocardiography (TEE) or epiaortic ultrasonography. The cannulation sites included the femoral artery, brachiocephalic artery, ascending aorta, and apex of the left ventricle.

The primary aim of this study was to compare the outcomes of antegrade perfusion via CAC and retrograde perfusion via FAC. Due to the small number of cases, patients who underwent BCAC or AC were excluded from the comparative analysis.

The primary outcomes were 30-day mortality and incidence of postoperative stroke. Organ malperfusion was defined on the basis of clinical symptoms and preoperative contrast-enhanced CT findings indicating compromised blood flow. Shock was defined as hypotension (systolic blood pressure 90 mmHg) caused by cardiac tamponade or left ventricular dysfunction, as demonstrated by transthoracic echocardiography, or the need for vasopressors to maintain blood pressure.

Stroke was categorized as dissection- or procedure-related. Dissection-related strokes were defined as preoperative neurological deficits such as impaired consciousness or hemiparesis. Procedure-related stroke was defined as new-onset stroke in patients without preoperative neurological symptoms. Stroke was included in both categories. When neurological abnormalities were observed, brain CT or magnetic resonance imaging was performed to diagnose cerebral infarction. However, asymptomatic cerebral infarctions cannot be diagnosed.

A standardized surgical protocol was applied to all patients, including rapid transfer to the operating room and immediate initiation of surgical repair [13, 14].

In all cases, except those involving apical cannulation, the ascending aorta was clamped before circulatory arrest and proximal anastomosis was performed. In patients undergoing hemiarch replacement (HAR) or TAR, the aorta was transected at the sinotubular junction level. The false lumen was obliterated using BioGlue® (CryoLife Inc, GA, USA), and the aortic wall was reinforced with inner and outer felt strips of polytetrafluoroethylene, secured with four 4-0 Prolene mattress sutures.

Aortic root replacement was performed in patients presenting annuloaortic ectasia or aortic root-entry tears. Patients with moderate to high aortic stenosis or regurgitation due to leaflet degeneration underwent valve replacement.

Distal aortic management was guided by preoperative CT or intraoperative findings. HAR was performed in cases where entry tears were confined to the ascending aorta, whereas TAR with a frozen elephant trunk (FET) was used in cases presenting arch or distal entry tears [15, 16]. In HAR, the aorta is transected just distal to the brachiocephalic artery and is reinforced with felt strips. In TAR, the aorta is divided into zone 2 just distal to the left common carotid artery. An FET was deployed, the outer layer was reinforced with felt, and the anastomosis was completed using a four-branch graft and continuous suturing with 4-0 polypropylene. The branches of the aortic arch were reconstructed using continuous 5-0 polypropylene continuous sutures.

Cerebral protection was achieved using deep hypothermic circulatory arrest (25 °C) and was monitored via bladder, rectal, and venous blood temperatures. In HAR procedures, retrograde cerebral perfusion was performed through the superior vena cava. In TAR procedures, selective antegrade cerebral perfusion was established after the aortic arch branches were transected.

The cannulation sites were selected based on the preoperative CT findings. Generally, FAC was the first choice. However, antegrade strategies (BCAC, CAC, or AC) have been used to minimize the embolic risk in patients with mural thrombus, aneurysmal changes, circumferential calcification, or a thrombosed false lumen descending to the abdominal aorta. Femoral cannulation was performed using the Seldinger technique with guidewire placement in the true lumen, as confirmed by TEE. For patients exhibiting limb malperfusion, perfusion was initiated on the ischemic side.

For BCAC procedures, the brachiocephalic artery was not dissected and direct cannulation (the so-called “Harutaro” technique) was performed after median sternotomy (Fig. 1, Ref. [17]). For CAC procedures, cannulation was performed at a site where the false lumen was thrombosed, and the aortic wall was sufficiently thick using the Seldinger technique under epiaortic echocardiographic guidance [18]. If no cannulation site was deemed safe, or in cases of preoperative cardiac arrest, apical cannulation was used. A purse-string suture with 4-0 polypropylene was placed at the apex, which was incised, and a cannula was inserted beyond the aortic valve.

In all cases, the absence of upper extremity hypotension and regional cerebral or lower limb oxygen desaturation was confirmed at the initiation of CPB. In one patient undergoing FAC, however, upper-extremity hypotension and cerebral desaturation developed after aortic cross-clamping and were resolved with supplementary inflow through the brachiocephalic artery.

All data were retrospectively collected from medical records. Categorical variables were compared using the chi-square test or Fisher’s exact test, and continuous variables were analyzed using the Student’s t-test. Data were expressed as medians and interquartile ranges and were compared using the Mann–Whitney U test. Multivariate logistic regression was used to identify independent predictors of 30-day mortality and postoperative stroke. Statistical significance was set at p 0.05.

All statistical analyses were performed using EZR v4.4.2 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R 4.0 (R Foundation for Statistical Computing, Vienna, Austria) [19].

Of the 92 patients who underwent surgery for ATAAD, FAC was performed in 71 patients (77%), CAC in 11 patients (12%), BCAC in six patients (7%), and AC in four patients (4%). The baseline characteristics of the patients are summarized in Table 1.

The preoperative clinical characteristics of the patients are shown in Table 2. In patients presenting with cardiopulmonary arrest, direct cannulation via the BCAC or the apex (AC) was preferred over Seldinger-based techniques such as FAC or CAC. A moderate or high aortic regurgitation (AR) was significantly more common in the CAC group than in the FAC group (Table 3).

All interventions performed on 82 patients included in the FAC vs. CAC analysis were performed by a single surgeon at the Fukuoka Wajiro Hospital. The operative data are summarized in Table 4. There were no significant differences in the proximal aortic procedures between the groups. However, TAR was significantly more frequent in the FAC group, likely accounting for the longer operative time observed in this group. Importantly, no significant difference in the time from skin incision to CPB establishment was observed between the two groups. The duration of circulatory arrest was also comparable between the groups.

The postoperative complications, including procedure-related stroke and 30-day mortality, are shown in Table 5. No statistically significant differences were observed between the groups for either outcome. The median postoperative hospital stay was 22 days in both groups.

Multivariate logistic regression analyses were performed to identify factors associated with procedure-related stroke and 30-day mortality. The factors related to procedure-related stroke are listed in Table 6. Variables such as age, sex, surgical procedure, presence of aortic arch branch dissection, FAC, CPB time, and circulatory arrest time were also examined. Dissection involving the aortic arch branches may pose a risk for procedure-related stroke; however, femoral artery cannulation was not associated with a significant increase in this risk.

As shown in Table 7, no significant associations were found between age, sex, preoperative malperfusion, and 30-day mortality. However, longer operative time was a significant predictor of 30-day mortality (odds ratio, 1.01; p = 0.00467).

The benefits of antegrade perfusion provided by CAC include prompt establishment of CPB after sternotomy, prevention of dissection propagation, avoidance of retrograde cerebral embolization and peripheral arterial injury, as well as improved neurological outcomes [21, 22]. However, few studies have concluded that CAC is associated with better outcomes in terms of stroke and mortality than FAC. Klotz et al. [7] compared 127 cases of CAC with 108 cases of FAC and reported new postoperative stroke rates of 8.7% in the central group and 5.6% in the femoral groups (p = 0.449). The 30-day mortality rates were 20.2% and 16.9%, respectively (p = 0.699). In the study by Klotz et al. [7], the cannulation site was selected at the discretion of the surgeon without any defined criteria.

Wang et al. [23] investigated cannulation strategies in DeBakey type I dissections, including AXC, FAC, and dual-arterial cannulation (DAC: a combination of AXC and FAC). The incidence of postoperative stroke was 6.0% in the FAC group and 2.7% in the AXC group (p = 0.02). Notably, DAC was associated with a higher stroke rate (6.9%) than AXC alone (2.5%; p 0.01). In-hospital mortality was significantly higher in the DAC group (14.4%) than in the AXC group (10.7%; p 0.01). These findings suggest that a single AXC scan is associated with the lowest risk of stroke. However, this study did not consider factors such as the vascular condition from the descending aorta to the femoral artery and the presence of thrombus in the false lumen, which are typical reasons for avoiding FAC. Therefore, FAC may have contributed to the increased incidence of postoperative stroke.

Recent studies have reported postoperative stroke rates ranging from 2.2% to 6% when an optimal cannulation strategy is employed [7, 18, 20, 23, 24]. In our study, the new-onset stroke rate in the FAC group was 2.8%, which compares favorably to previous reports. This may reflect our institutional practice of avoiding FAC in patients with poor distal aortic morphology or in those who are at a high risk of thromboembolism.

Liang et al. [24] also found no significant reduction in stroke or 30-day mortality between DAC and single-site cannulation. However, in a subgroup analysis of patients with CPB times exceeding 180 min, DAC was associated with improved survival compared to both axillary (HR: 0.35, 95% CI: 0.14–0.90, p = 0.029) and femoral cannulation (HR: 0.38, 95% CI: 0.14–1.03, p = 0.058). Although DAC interventions are considered to provide adequate systemic perfusion, reports on their effectiveness are limited [25, 26]. In our study, none of the patients experienced a CPB time of $>$180 min. Therefore, we considered that it was not appropriate to evaluate the utility of DAC. The median CPB time was 93 min in the FAC group and 80 min in the CAC group, both of which were relatively short. This may have contributed to the low mortality rate and limited benefits observed with DAC. Furthermore, when CPB time is short, prolonged overall operative time may increase 30-day mortality.

Axillary or brachiocephalic cannulation often involves graft interposition, which can delay the initiation of CPB. This delay may increase the risk of malperfusion and aortic rupture. Norton et al. [27] reported median times from skin incision to CPB of 41 min for CAC and 85 min for arch vessel cannulation, with similar stroke (5.6% vs. 5.2 %) and mortality rates.

Patients with cardiac tamponade more often required arch vessel cannulation, likely because CAC is difficult to perform during hemodynamic instability. In these cases, gradual tamponade decompression followed by FAC or AXC is usually preferred. In our cohort, 45% of patients undergoing CAC presented with cardiac tamponade, and cardiopulmonary bypass was not started via peripheral cannulation beforehand. Instead, careful blood pressure control by the anesthesiologist and close teamwork with the surgical staff were crucial. At our institution, direct brachiocephalic cannulation, referred to as ‘Harutaro’ cannulation, is performed in selected cases where the arch vessels are free of dissection [17]. Additionally, recent reports have described the “Samurai” technique, in which the ascending aorta is directly incised and cannulated into the true lumen. However, this approach did not demonstrate significant improvements in mortality or stroke rates [28, 29].

Postoperative stroke risk is influenced by a complex interplay of anatomical and procedural factors, including vessel morphology, entry and re-entry sites, and the presence of a thrombosed false lumen. A more detailed understanding of these elements may help to refine cannulation strategies to improve patient-specific outcomes.