Evaluation of remodeling of the descending aorta using 3DCT after total arch replacement for acute aortic dissection Authors: Manato Saitoh, Tomohiro Imazuru, Masateru Uchiyama, Tomoki Shimokawa Published: 2025 ABSTRACT AbstractObjectivesThe frozen elephant trunk (FET) technique during total arch replacement (TAR) in patients with acute type A aortic dissection (ATAAD) has been shown to promote favorable aortic remodeling. However, few reports have compared the morphological assessments between the conventional elephant trunk (CET) and FET procedures. Here, we investigated whether the FET technique during TAR for the patients with ATAAD could affect aortic remodeling through morphological analysis.MethodsIn total, 464 patients diagnosed with ATAAD underwent emergency ascending aorta or TAR between 2010 and 2023. We retrospectively analyzed the clinical data of 28 patients who underwent TAR and postoperative contrast-enhanced computed tomography. We assigned 28 patients to either the CET group (n = 11) or FET group (n = 17). Patient characteristics, postoperative outcomes, and aortic remodeling analysis were collected.ResultsAortic remodeling analysis showed that the expansion rate of the true lumen (TL) in the descending thoracic aorta was significantly greater in the FET group early and 1 year after surgery. Meanwhile, the shrinkage rate of the false lumen (FL) in the descending thoracic aorta was significantly greater in the FET group early after surgery and tended to be greater in the FET group even 1 year after surgery. Remarkably, the incidence of reintervention of the descending aorta was significantly lower in the FET group.ConclusionsThe FET technique with TAR in patients with ATAAD could improve aortic remodeling through improved TL expansion and preventive FL dilatation in the descending thoracic aorta and reduce the need for surgical reintervention.Keywords: Acute type A aortic dissection, Total arch replacement, Frozen elephant trunk, Aortic remodeling Introduction Progress has paved the way for the surgical outcomes of acute type A aortic dissection (ATAAD) because of the progress in surgical techniques and medical devices. Surgical strategies considering long-term outcomes have started to be adopted at the initial intervention for ATAAD [1, 2]. In particular, studies have shown that total arch replacement (TAR) is effective in patients with primary entry located beyond the aortic arch and those with extensive tears beyond the aortic arch branches [3–5]. However, the potential risk of residual false lumen (FL) enlargement in the descending aorta should be considered during long-term follow-up after TAR in patients with ATAAD, which may require additional treatment. Recently, the implantation of a frozen elephant trunk (FET) prosthesis during TAR has been shown to promote distal FL thrombosis and improve long-term outcomes, including beneficial remodeling of the descending aorta, reducing future complications [6, 7]. Nevertheless, few reports have compared the remodeling of the descending aorta between the conventional elephant trunk (CET) and FET procedures for TAR. This study aimed to morphologically evaluate whether the FET technique for TAR in patients with ATAAD could affect the aortic remodeling of the descending and abdominal aorta compared with the CET technique using three-dimensional computed tomography (3DCT). Methods Study design A total of 464 patients diagnosed with ATAAD with patent FL underwent emergency ascending aorta or TAR at our institution between January 2010 and March 2023. Of these, 109 underwent TAR, and their clinical characteristics were retrospectively reviewed and analyzed. Of the 109 patients, we assigned 28 patients, who had available medical records including contrast-enhanced 3DCT scans, to undergo either the CET procedure (CET group; n = 11) or FET procedure for TAR (FET group; n = 17) (Fig. 1). The remaining 81 patients were excluded for the following reasons: (1) in-hospital deaths (12 cases), (2) renal dysfunction or allergies to contrast agents (47 cases), (3) insufficient CT evaluation because of withdrawal (19 cases), (4) lack of follow-up (3 cases). Postoperative contrast-enhanced 3DCT after discharge was performed early and 1 year after surgery to evaluate the volumes of the true lumen (TL) and FLs in the descending aorta and the rate of change in each volume. CT scans taken at 1 mm slices were semi-automatically converted to three-dimensional data using a workstation (Ziostation2 Version 2.4.0.4, Ziosoft Inc., Tokyo, Japan), and manual corrections were performed to measure the volume. All measurements were performed by an unrelated researcher. Enrollment. CET, conventional elephant trunk; CT, computed tomography; FET, frozen elephant trunk; TAR, total arch replacement Device and its selection After the FET prosthesis became commercially available as the Frozenix J Graft (Japan Lifeline Co., Ltd, Tokyo, Japan), it became the first choice for patients who underwent TAR after 2015. CET procedure was performed for all cases before 2015, and FET procedure was performed for all cases after 2015. Frozenix consists of a distal stented part with oval-shaped nitinol stents and a proximal non-stented graft. The total length of the Frozenix is 200 mm, and the stented part can be 60, 90, or 120 mm long. After identifying the dissection entry location and measuring the descending aorta diameter using preoperative contrast-enhanced 3DCT, the surgeons selected a graft diameter in increments of 2 mm (21–39 mm) to be approximately 90% of the descending aorta diameter as the planned distal landing site. Additionally, the length of the stented part was decided not to exceed Th8 in the straight portion of the descending aorta to decrease the risk of paraplegia. Surgical and endovascular procedures According to our institutional policy, all patients underwent emergency surgery, with no delay between diagnosis and surgery. All procedures were performed via a median sternotomy. Standard cardiopulmonary bypass was performed using the right or left common femoral arterial and right atrial cannulations. In cases of severe calcification and dissection of the common femoral arteries, arterial cannulation was achieved through the axillary arteries or apex. In all patients, open distal anastomosis was performed under moderate hypothermic circulatory arrest at a rectal temperature of 25 °C and antegrade selective cerebral perfusion at a rate of 8–12 mL/kg/min to the innominate, the left common carotid, and the left subclavian arteries through arteriectomy. In the CET group, distal anastomosis was constructed using a ‘stepwise anastomosis technique’ with the elephant trunk procedure [8]. In the FET group, a distal anastomosis was constructed between the left common carotid and left subclavian arteries (Zone 2). The left subclavian artery was closed at the proximal stump. The Frozenix was inserted and deployed in the TL under transient retrograde distal perfusion through the femoral artery or transient antegrade perfusion through a balloon catheter inserted from the aortic arch transection site to the descending aorta to prevent air embolism of the intercostal arteries. After reinforcing the distal stump of the aorta with a Teflon-felt strip, the quadrifurcated prosthesis was anastomosed to the stump of the descending aorta. Rewarming was initiated after cross-clamping the proximal graft and antegrade systemic circulation from the fourth limb. The proximal graft was anastomosed to the ascending aorta, reinforced with Teflon-felt strips on the inside and outside after fixation of the dissected aortic wall with BioGlue (Cryolife, Kennesaw, USA). The reconstruction of the aortic arch branches was then reconstructed. Follow-up data collection and aortic remodeling analysis All patients without renal dysfunction or allergies were evaluated using contrast-enhanced 3DCT imaging during hospitalization and 1 year after surgery at our institution or other related hospitals. After multiplanar reconstruction of the preoperative and postoperative CT image using a workstation (Ziostation2), the volumes of the TL and FL from the carina to the bifurcation of the celiac trunk (descending thoracic aorta) and from the bifurcation of the celiac trunk to the terminal aorta (abdominal aorta) were measured to calculate the ratio of volume change early and 1 year after surgery. The ratio of the volume change at each follow-up period was calculated as follows: Statistical analysis Our data were summarized by calculating means and standard deviations for normally distributed variables, medians and interquartile ranges (IQR) for non-normally distributed variables, and frequencies and percentages for categorical variables. Bivariate differences were compared using the unpaired t test or Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. Statistical significance was set at p < 0.05. All statistical analyses were performed using JMP Pro 17.0 for Windows (SAS Institute, Cary, NC, USA). Ethical approval All subjects enrolled in this research gave their informed consent, which alongside the study design, has been approved by the institutional committee on human research (24–064). Results Preoperative and intraoperative data The preoperative patient details of both groups are listed in Table 1. There were no significant differences in any of the variables between the groups. Table 1Patient characteristicsOverall (n = 28)CET (n = 11)FET (n = 17)p valueAge (years)53.7 ± 7.653.9 ± 6.753.5 ± 8.30.900Female gender4 (14.3)1 (9.1)3 (17.7)0.528Body surface area (m2)1.79 ± 0.191.80 ± 0.111.78 ± 0.230.821Body mass index (kg/m2)24.2 ± 4.223.3 ± 3.324.7 ± 4.70.419Past medical history Hypertension22 (78.6)9 (81.8)13 (76.5)0.736 Dyslipidemia3 (10.7)1 (9.1)2 (11.8)0.823 Chronic kidney disease10 (35.7)3 (27.3)7 (41.2)0.453 Coronary artery disease1 (3.6)0 (0.0)1 (5.9)0.413Type of aortic dissection DeBakey type I17 (60.7)5 (45.5)12 (70.6)0.184 Retrograde dissection11 (39.3)6 (54.6)5 (29.4)0.184Organ malperfusion Neck vessels0 (0.0)0 (0.0)0 (0.0)- Coronary1 (3.6)0 (0.0)1 (5.9)0.413 Visceral1 (3.6)0 (0.0)1 (5.9)0.413Data are presented as means ± standard deviations or n (%) unless noted otherwiseCET conventional elephant trunk, FET frozen elephant trunk Patient characteristics Data are presented as means ± standard deviations or n (%) unless noted otherwise CET conventional elephant trunk, FET frozen elephant trunk Both the FET and CET grafts were successfully deployed or inserted in all patients. Intraoperative variables of both groups are shown in Table 2. Regarding the location of primary entry, the aortic arch was more prevalent in the CET group than in the FET group (p = 0.014). The mean operation time (385.0 ± 132.6 min vs. 406.6 ± 95.3 min, for CET and FET groups, respectively, p = 0.620), cardiopulmonary bypass time (206.4 ± 53.3 min vs. 230.8 ± 56.2 min, for CET and FET groups, respectively, p = 0.261), aortic cross-clamp time (175.5 ± 45.5 min vs. 181.8 ± 38.3 min, for CET and FET groups, respectively, p = 0.693), circulatory arrest time (54.5 ± 21.5 min vs. 42.5 ± 15.3 min, for CET and FET groups, respectively, p = 0.095), and selective cerebral perfusion time (105.7 ± 27.2 min vs. 104.2 ± 36.9 min, for CET and FET groups, respectively, p = 0.906) were not significantly different. Table 2Intraoperative characteristicsOverall (n = 28)CET (n = 11)FET (n = 17)p valueOperation time (min)398.1 ± 109.6385.0 ± 132.6406.6 ± 95.30.620Cardiopulmonary bypass time (min)221.2 ± 55.4206.4 ± 53.3230.8 ± 56.20.261Aortic cross-clamp (min)179.3 ± 40.6175.5 ± 45.5181.8 ± 38.30.693Circulatory arrest time (min)47.3 ± 18.654.5 ± 21.542.5 ± 15.30.095Selective cerebral perfusion time (min)104.8 ± 32.9105.7 ± 27.2104.2 ± 36.90.906Arterial cannulation Femoral artery22 (78.6)9 (81.8)13 (76.5)0.736 Ascending aorta3 (10.7)1 (9.1)2 (11.8)0.823 Apex3 (10.7)1 (9.1)2 (11.8)0.823Location of primary entry Ascending aorta14 (50.0)3 (27.3)11 (64.7)0.053 Aortic arch8 (28.6)6 (54.6)2 (11.8)0.014 Descending aorta6 (21.4)2 (18.2)4 (23.5)0.736Primary entry resection28 (100.0)11 (100.0)17 (100.0)1.000Length of FET 60 mm––1 (5.9)– 90 mm––9 (52.9)– 120 mm––7 (41.2)–Concomitant procedures Coronary artery bypass grafting3 (10.7)1 (9.1)2 (11.8)0.823 Root replacement1 (3.6)0 (0.0)1 (5.9)0.413Data are presented as means ± standard deviations or n (%) unless noted otherwiseCET conventional elephant trunk, FET frozen elephant trunk Intraoperative characteristics Data are presented as means ± standard deviations or n (%) unless noted otherwise CET conventional elephant trunk, FET frozen elephant trunk Postoperative clinical outcomes The postoperative variables in both groups are listed in Table 3. The intensive care unit stay and length of hospital stay revealed no statistically significant differences between the CET and FET groups (6 [IQR 4–9] days vs. 3 [3–4.5] days and 20 [13–21] days vs. 15 [11–40.5] days, respectively). No 30-day mortality was noted; however, the CET group tended to have a higher all-cause mortality than the FET group (18.2% vs. 0%, p = 0.068). The causes of all-cause mortality in the CET group were sepsis owing to urinary tract infection (n = 1) and myocardial infarction (n = 1). Remarkably, the incidence of reintervention in the descending aorta was significantly higher in the CET group than in the FET group (27.3% vs. 0%, p = 0.023). Reintervention in the descending aorta was performed in three patients with enlargement of the descending aorta in the CET group. One patient underwent thoracic endovascular aortic repair one and a half years after surgery, and two patients underwent descending aorta replacement 6 and 9 years after surgery. The incidences of other major complications did not differ. Table 3Postoperative clinical outcomesOverall (n = 28)CET (n = 11)FET (n = 17)p valueICU length of stay (day), median (IQR)4 (3–6.8)6 (4–9)3 (3–4.5)0.068Hospitalization (day), median (IQR)16.5 (11.3–25.8)20 (13–21)15 (11–40.5)0.60430-day mortality0 (0.0)0 (0.0)0 (0.0)–MACCE All-cause mortality2 (7.1)2 (18.2)0 (0)0.068 Cardiac death1 (3.6)1 (9.1)0 (0)0.206 Vascular death0 (0)0 (0)0 (0)– Myocardial infarction2 (7.1)2 (18.2)0 (0)0.068 Reintervention for descending aorta3 (10.7)3 (27.3)0 (0)0.023 Stroke2 (7.1)2 (18.2)0 (0)0.068Deep wound infection5 (17.9)2 (18.2)3 (17.7)0.971Hemodialysis1 (3.6)0 (0)1 (5.9)0.413Distal SINE1 (3.6)0 (0)1 (5.9)0.413Paraplegia1 (3.6)0 (0)1 (5.9)0.413Paraparesis2 (7.1)1 (9.1)1 (5.9)0.748Data are presented as median (interquartile range) or n (%) unless noted otherwiseICU intensive care unit, IQR interquartile rage, SINE stent graft-induced new entry tear Postoperative clinical outcomes Data are presented as median (interquartile range) or n (%) unless noted otherwise ICU intensive care unit, IQR interquartile rage, SINE stent graft-induced new entry tear Volume remodeling at the downstream aorta Descending thoracic aorta After examining the chronological change ratio in the volume from the level of the carina to the level of the bifurcation of the celiac trunk, aortic remodeling analysis showed that the expansion rate of the TL in the descending thoracic aorta was significantly greater in the FET group than in the CET group early and 1 year after surgery (Fig. 2A). Additionally, the shrinkage rate of the FL in the descending thoracic aorta was significantly greater in the FET group than in the CET group early after surgery and tended to be greater in the FET group even 1 year after surgery (Fig. 2B). Volume remodeling at the downstream aorta. Chronological change ratio in the volume of A true lumen (TL) and B false lumen (FL) at the descending thoracic aorta and C TL and D FL at the abdominal aorta. * p < 0.05 compared with the CET group. CET, conventional elephant trunk; FET, frozen elephant trunk; FL, false lumen; TL, true lumen Abdominal aorta After examining the chronological change ratio in the volume from the bifurcation of the celiac trunk to the terminal aorta, aortic remodeling analysis showed no significant change in the ratio of the TL and FL in the abdominal aorta between the groups during the follow-up period; however, the TL tended to be greater in the FET group than in the CET group, but the FL tended not to be greater in the FET group than in the CET group (Fig. 2C, D). Discussion In the 2021 annual report of the Japanese Society of Thoracic Surgery, the number of surgeries for acute aortic dissection was 6416, which was the highest number ever recorded [9]. Regarding the choice of surgical procedure, TAR was performed in 2092 cases (32.6%) and ascending replacement in 1934 cases (30.1%). Additionally, the number of TAR with FET has increased annually, with 1363 cases (21.5%), 1648 cases (26.6%), and 1876 cases (29.2%) in 2019, 2020, and 2021, respectively [9]. In particular, as a therapeutic strategy, the efficacy of FET for ATAAD in aortic remodeling and aortic-related events has been reported [10]; however, any conclusion on this issue remains speculative, as these studies lacked morphological evaluation for postoperative image analysis. In numerous cases, evaluations have relied on two-dimensional assessments of specific cross-sections of CT imaging. This method can lead to inaccuracies when comparing individual cross-sections taken at different times. To enhance precision and ensure a thorough chronological comparison, it is essential to implement three-dimensional evaluations. Furthermore, there has been no consensus thus far on whether some variables such as aortic diameter, aortic cross-sectional area, and thrombosis rate of the FL should be considered reliable for evaluating descending aortic remodeling in the long-term after ATAAD. Here, we report the first study in Japan to investigate aortic remodeling by calculating the chronological change ratio and volume of the TL and FL using preoperative and postoperative 3DCT imaging after TAR with CET or FET for ATAAD. The most remarkable finding of our study is that TAR with the FET technique was associated with significantly more effective expansion of the TL and prevention of FL dilatation at the descending thoracic aorta than with CET in patients with ATAAD; however, aortic remodeling in the abdominal aorta was not observed. Generally, Frozenix is known as a self-expanding stent graft with a radial force of approximately 0.9–1.8 kPa and can facilitate the covering of the primary entry tears with its functional radial force and appropriate attachment. It is suggested that these aspects result in a reliable higher rate of complete FL thrombosis (96.8%), beneficial aortic remodeling, and promising improved long-term prognosis in the descending thoracic aorta of patients with ATAAD who underwent TAR with FET [11, 12]. Additionally, previous studies, mainly based on the analysis using 4D Flow magnetic resonance imaging after the FET technique in patients with ATAAD, have described that the almost complete disappearance of antegrade blood flow through the FL at the FET landing zone promoted thrombosis of the FL, which could sequentially contribute to favorable remodeling [13–16]. In a recent research addressing thrombosis rate and remodeling of the FL of the descending aorta after FET, the descending aorta after TAR with FET was assessed in three specific regions: (1) the stent-graft landing site, (2) from the distal end of the stent-graft to the level of the celiac artery, and (3) from the celiac artery to the terminal aorta. This study indicated that postoperative thrombosis in the FL decreased as the distance from the stent-graft landing site increased, and the remodeling effect of the FET tended to diminish [17]. Additionally, retrograde blood flow was observed to remain flowing through the FL via some re-entries occurring in the distal area from the stent-graft landing site in the descending aorta with the FET procedure; this may delay the complete FL thrombosis and induce negative remodeling [14]. Although direct comparison was not possible, the effect of aortic remodeling by FET was significantly preserved in the descending thoracic aorta, even though the FET group had more primary entry in the descending aorta as a preoperative background. Consistent with previous studies and our results, Kozlov et al. have reported that the group using a long FET device achieved better aortic remodeling than the group using a short FET device [18]. In light of these findings, it appears that an extended FET landing zone is likely to enhance the closure of re-entry, promoting more significant and effective outcomes in aortic remodeling. Nevertheless, there is still debate regarding the optimal length of FET devices, since the use of excessively long FET devices increases the risk of spinal cord ischemia associated with the FET technique, as has been already reported [19]. Although further evidence on the optimum length of the FET is necessary, our data indicated that volumetric evaluation using detailed 3DCT images revealed that using a FET with a length not exceeding Th8 resulted in favorable remodeling in the descending thoracic aorta without significantly increasing complications such as spinal cord ischemia. Our strategy on the choice of FET length could not be less effective for FL of the abdominal aorta with retrograde blood flow through potential residual re-entries. There is a concern regarding the incidence of reintervention of the descending aorta after TAR with FET in patients with ATAAD. Generally, a patent or incompletely thrombosed FL at the descending thoracic aorta after initial ATAAD surgery is thought to be a strong predictor of distal revascularization and is associated with poor prognosis [20, 21]. A recent study has reported that the postoperative FL patency in patients after surgery for ATAAD was 62%. Moreover, 94% of patients who died of aortic rupture developed a patent FL. The incidence of distal reoperation in patients with ATAAD after the initial surgery was 9.5% and tend to be higher in the descending thoracic aorta [20]. Collectively, the presence of patent FL after surgery for ATAAD could be a significant risk factor for sudden death and aortic-related complications, leading to poor long-term survival [22]. In this study, the FET group achieved significantly better aortic remodeling in the descending thoracic aorta and showed a higher trend of aortic remodeling in the abdominal aorta. At up to approximately 6 years of follow-up, no additional surgical reintervention of the residual descending aorta was required in the FET group. However, 27.3% of patients in the CET group required reintervention (p = 0.023). To this point, there is limited evidence to determine whether aortic remodeling by FET could improve the prognosis. However, our findings indicate that the FET technique at initial surgery can potentially mitigate the risk of reintervention of the descending aorta through effective expansion of the TL and prevention of FL dilatation. Study limitations This study has several limitations that should be considered when interpreting its findings. First, the sample size is notably small, with only 28 patients included over a 1-year period. The limited number of participants reduced the statistical power of the study and may not have provided a comprehensive representation of the broader population of patients with ATAAD who underwent TAR. Second, the non-randomized and retrospective design of this study introduced several biases, such as selection bias, which could affect the validity of the outcomes. Notably, the difference in data collection timing between the CET and FET groups was carefully considered, depending on the specificities and maturity of our facility. Many patients in the CET group were enrolled in the first half of the investigation period, whereas the patients in the FET group were enrolled in the second half of the investigation period. Finally, the 1-year follow-up did not provide a promising overview of the potential benefit of the FET technique on aortic remodeling, as FL dilatation and aneurysms in the descending thoracic aorta can develop slowly. Resolving the problems of many facilities evaluating aortic volumetric remodeling using 3DCT may allow for further investigations. Conclusion The FET technique with TAR resulted in more favorable postoperative aortic remodeling than CET through improved TL expansion and preventive FL dilatation in the descending thoracic aorta in patients with ATAAD. These favorable effects are expected to reduce the incidence of reintervention associated with the dilatation of the residual dissected aorta. Acknowledgements We would like to thank Editage (www.editage.com) for English language editing. Author contributions Manato Saitoh—designing the study, collecting the data, analyzing data and images, writing article, reviewing patient notes, and approving final submission. Tomohiro Imazuru—corresponding author; designing the study, analyzing data and images, writing article, reviewing patient notes, and approving final submission. Masateru Uchiyama—designing the study, collecting the data, analyzing data and images, writing article, reviewing patient notes, and approving final submission. Tomoki Shimokawa—reviewing patient notes, analyzing data and images, and approving final submission. Funding None. Data availability The datasets used are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate All patients provided informed consent, and the study protocol was approved by the Ethics Committee of Teikyo University. Consent for publication Written consent was obtained from all patients for writing this study and accompanying data. Identifying details have been omitted. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used are available from the corresponding author on reasonable request.