The soft elephant trunk: a new approach in the treatment of aortic dissection

Introduction

For 30 years, the hybrid approach to management of aortic dissection (AD) has been actively evolving. Despite progress of the frozen elephant trunk (FET) procedure, complications such as distal stent-graft-induced new entry (dSINE), negative remodeling and need for reoperation remain significant challenges. Conventional hybrid prostheses [E-Vita Open Plus (Jotec, Hechingen, Germany), Thoraflex Hybrid (Terumo, Inchinnan, Scotland), etc.] have limitations in adapting to dissected aortic anatomy, and a high rate of distal stent-graft related complications. This study evaluates the long-term effectiveness of the “Soft Elephant Trunk” (SET) hybrid prosthesis (MedInzh, Penza, Russia) in reducing these complications and improving clinical outcomes.

Methods

This single-center retrospective cohort study included consecutive patients with AD treated between 2014 and 2024. Inclusion criteria were: (I) any type of AD; (II) age >18 years; (III) hybrid repair of AD propagating distal to the left subclavian artery; and (IV) prosthesis selection independent of patient-specific anatomy. Aortic dissection diagnosis was confirmed by computed tomography (CT) in all patients.

Study design

To evaluate the efficacy of the dissection-specific (DS) hybrid aortic prosthesis, patients were divided into two groups: DS group and non-DS group. To minimize selection bias, equal groups were formed using propensity score matching. The study was approved by the Local Ethics Committee of Petrovsky National Research Centre of Surgery (protocol#7 15.04.2021; #1 24.01.2025) and written informed consent was waived. The primary endpoint was any new distal aortic event. Secondary endpoints were mortality and postoperative complications. All patients underwent sequential CT scans according to the protocol. To ensure complete imaging data for the assessment of mid-term aortic remodeling, only those subjects who completed every scan at the specified intervals—preoperatively, during the hospital stay, at 6 months, at 1 year, and annually thereafter—with a minimum follow-up of 5 years were included in the analysis.

Aortic remodeling assessment

Aortic remodeling was assessed using an institutional protocol specifically developed for patients with AD. Planimetric analysis of the true lumen (TL), false lumen (FL), and total aortic lumen (AL) areas and perimeters was conducted at seven predefined aortic levels on all serial CT scans. Volumetric changes were evaluated at three levels: within the stent-graft, distal descending thoracic aorta (DTA), and abdominal aorta. The status of the FL (patent, partially thrombosed, or completely thrombosed) was assessed at five anatomical levels: within the stent-graft, distal DTA, level of visceral arteries, abdominal aorta, and iliac arteries.

Evolution of surgical technique

In our center, we utilize various cannulation options depending on AD anatomy and clinical context, though the right subclavian artery is preferred in most cases. The brain protection strategy relies on antegrade cerebral perfusion, typically bilateral. Neuromonitoring includes transcranial Doppler and cerebral oximetry. All surgeries are performed under moderate hypothermia (1). Temperature is routinely monitored at two sites: the urinary bladder and the nasopharynx. Distal first technique is used, beginning with aortic arch reconstruction and distal anastomosis, followed by stent-graft implantation. Distal perfusion is initiated after anastomosis with the left subclavian artery (multibranched prostheses) or after anastomosis with the brachiocephalic vessels (island technique). A guidewire was initially used for stent-graft implantation, but it is now reserved for complex cases. In all cases, we avoided excessive oversizing at the distal edge, especially in high-risk patients. All procedures were performed by a single surgeon. Prosthesis sizing was based on intraoperative TL measurement at zone 3 using a sizer, and on preoperative CT-based assessment at the level of the left atrium, guided by the TL perimeter.

• d TL = circumference length/π
• Acute AD: d SG = d TL + 10% (oversizing)
• Subacute AD: d SG = d TL (+10% oversizing in cases of malperfusion)
• Chronic AD: d SG = d TL.

In cases of narrow TL, we performed a wedge resection of the intima at the distal anastomosis, thereby eliminating the acute angle at the anastomosis. Before 2014, we primarily employed a hybrid technique that included the Elephant Trunk procedure with following TEVAR. We started using hybrid prostheses in our center in 2014, with the first prostheses being Thoraflex Hybrid and E-Vita Open Plus. Since 2019, we have introduced our own DS hybrid prosthesis, SET (MedEng, Penza, Russia) (2,3). The concept of SET is the progressive reduction of radial stiffness in the endovascular part. This is primarily achieved by decreasing the stiffness of the nitinol rings of the stent graft towards the distal end through a proportional reduction in the thickness of the nitinol rings. The final ring has been modified from an S-shape to a Z-shape, with its stiffness reduced by 50% compared to the first ring. Secondly, the distal end features a 2 cm vascular graft with nine platinum markers. This design prevents the distal end from exerting pressure on the fragile intima, thereby preventing dSINE. The third important advantage is the length of the stent graft. The hybrid graft measures 200 mm in length and lacks a suture ring. This design feature, combined with a compact delivery system, facilitates implantation below Th8, effectively preventing distal fenestrations (Figure 1).

Figure 1 Soft elephant trunk. (A) dissection-specific “Soft Elephant Trunk” prosthesis (MedEng, Penza, Russia); (B) patient with acute Stanford type A aortic dissection before and after FET procedure using a DS stent graft (“SET”); (C) patient with chronic Stanford type non-A-non-B aortic dissection before and after FET procedure using a DS stent graft (“SET”). DS, dissection-specific; FET, frozen elephant trunk; SET, soft elephant trunk.

Statistical methods

Data analysis was conducted using parametric or nonparametric methods, depending on variable distribution. Quantitative variables are presented as mean (standard deviation) for normal distribution and median (IQR) for non-normal distribution. Qualitative variables are expressed as counts and percentages. The Shapiro-Wilk test assessed normality. Student’s t-test was used for normally distributed continuous variables, with the Mann-Whitney U test applied otherwise. Categorical variables were compared using Fisher’s exact test and Chi-squared test. Paired statistical tests were used where applicable, especially for comparisons between post-matching groups (McNemar’s test quantitative variables, paired t-test or the Wilcoxon signed-rank test for qualitative variables). Propensity score matching (PSM) was performed using a Greedy algorithm with Euclidean distance to match treated units with control units on a one-to-one basis. A 95% confidence interval was used to ensure reliable estimates, and a strict tolerance threshold of 0.001 was applied to match units with highly similar propensity scores. We used sex, body mass index (BMI), AD type and phase, DTA dilation, prior aortic interventions, and CTD as covariates for PSM, which resulted in 71 matched pairs. The Kaplan-Meier method and log-rank test were used to analyze survival rate, aortic event-free rates, and negative remodeling. Multivariable Cox regression was used to identify the predictors of dSINE. Statistical analysis was performed using Jamovi version 2.3.21.0. PSM was performed using XLSTAT version 2024.4.1. A P value of less than 0.05 was considered statistically significant.

Results

Demographic data

A total of 323 patients underwent FET (2014–2024). Of these, 241 cases of AD were included. Patients were divided into two groups: DS group (n=170) and non-DS group: (n=71). Median age was 53 [45, 63] vs. 52 [45, 62] years (P=0.6), respectively. There were no significant differences in sex, BMI, connective tissue disorder (CTD), hypertension, etc. between groups. A higher proportion of type A AD was observed in DS group: 122 (72%) vs. 63 (89%) (P=0.017), but after matching, this difference was non-significant. DS group had a significantly higher prevalence of aortic arch aneurysms (P=0.005 and 0.002 in both comparisons). DS group had a significantly higher rate of chronic kidney disease [51 (30%) vs. 9 (13%), P=0.005], and patients with malperfusion, which disappeared after matching (P>0.05) (Table 1).

Perioperative outcomes

The DS group typically had smaller grafts (26 mm; IQR: 24, 28 mm) compared to non-DS group, who were more frequently fitted with larger grafts (28 mm; IQR: 28, 30 mm) (P<0.001). Additionally, DS group had a longer graft length on average (P<0.001). DS group had significantly shorter CPB, aortic cross-clamp, circulatory arrest times, and hypothermia compared to non-DS patients, which we associate with the learning curve (Table 2). The incidence of pulmonary failure, stroke, and cardiac arrest was comparable across groups. Overall hospital mortality was similar between groups, at 5.3% in DS versus 7% in non-DS (P=0.56) and 5.6% versus 7% after matching (P>0.99) (Table 3).

Long-term outcomes

Late survival

We evaluated long-term outcomes in both overall and PSM groups. The follow-up duration significantly differed between the non-DS and DS groups [18 (7.9, 40) vs. 60 (34.3, 85.8) months, P<0.01], including after PSM adjustment [18 (9.2, 42.8) vs. 60 (34.3, 85.8) months, P<0.01]. In the DS group, late mortality (beyond 30 days) due to aortic-related causes occurred in three cases (1.7%): one case of aortic rupture in the visceral segment after TAAA replacement in a patient with Loeys-Dietz syndrome and two cases of aortic graft infection. In the non-DS group, late mortality due to aortic-related causes occurred in seven cases (9.8%): one case with dSINE complicated by distal aortic rupture, five cases of aortic graft infection, and one case of proximal aortic rupture. Survival rate was 85.4% (95% CI: 77.5–94.2%) vs. 90.5% (95% CI: 85.5–95.9%) and 78.5% (95% CI: 69–89.2%) vs. 76.5% (95% CI: 58–100%) at 3 and 5 years, in non-DS and DS groups (P=0.36), respectively (Figure 2A). In the matched cohort, survival rate was 85.4% (95% CI: 77.5–94.2%) vs. 92.3 (95% CI: 84.8–100%) and 78.5% (95% CI: 69–89.2%) vs. 73.9% (95% CI: 47.3–100%) at 3 and 5 years, in non-DS and DS groups (P=0.24), respectively (Figure 2B).

Figure 2 Follow-up outcomes stratified by DS and non-DS groups. Kaplan-Meier curves for (A) overall survival (non-matched cohort), (B) overall survival (matched cohort), (C) non-staged distal aortic reoperations (non-matched cohort), (D) non-staged distal aortic reoperations (matched cohort), (E) freedom from dSINE (non-matched cohort), (F) freedom from dSINE (matched cohort). DS, dissection-specific; dSINE, distal stent-graft induced new entry.

Late reoperations

Overall reoperation rates were significantly lower in the DS group compared to the non-DS group (10% vs. 24%, P=0.005), although this difference became non-significant after propensity matching (P=0.13). Non-staged reoperations were more frequent in the non-DS group (18% vs. 4.7%, P<0.001), while staged reoperations showed no significant difference between groups (P>0.99). After performing PSM analysis, the DS group demonstrated a significantly lower frequency of non-staged reoperations compared to the non-DS group (5.6 vs. 18%, P=0.02). Both groups had comparable frequencies and types of distal reoperations (Table 3); however, after PSM, there was a trend towards more frequent TEVAR procedures in the non-DS group (9.9% vs. 4.2%, P=0.326) and more frequent open surgeries in the DS group (11.3% vs. 9.9%, P=1.000). This shift is attributed to the evolving surgical strategy. These findings indicate that the DS group had a lower reoperation burden, particularly for non-staged procedures, but after adjusting for baseline differences, most comparisons lost statistical significance.

Freedom from reoperation was 89.5% (95% CI: 81.9–97.9%) vs. 91.3% (95% CI: 85–98%) and 86.56% (95% CI: 77.5–96.7%) vs. 91.3% (95% CI: 85–98%) at 3 and 5 years, in non-DS and DS groups (P=0.5), respectively (Figure 2C). In the matched cohort, freedom from reoperation was 89.5% (95% CI: 81.9–97.9%) vs. 91% (95% CI: 82.8–100%) and 86.56% (95% CI: 77.5–96.7%) vs. 91% (95% CI: 82.8–100%) at 3 and 5 years, in non-DS and DS groups (P=0.78), respectively (Figure 2D).

Aortic remodeling

To assess aortic remodeling, we selected 106 patients [non-DS (n=50) and DS (n=56) (Table S1)]. Type A AD was more prevalent in both groups but slightly lower in DS group (73.2% vs. 86.0%, P=0.105). Chronic AD was more frequent in DS group (62.5% vs. 54.0%), and non-DS group had a higher proportion of subacute AD (28.0% vs. 14.3%, P=0.216). Aortic diameters were measured at three key levels [stent-graft level (A), distal DTA (B), and abdominal aorta (C)] with no statistically significant differences between groups (P>0.05), and the ratios of TL/AL also remained comparable (Figure S1A). The number of re-entries at various levels (A, B, C) was higher in DS group, but significance was only reached for re-entries >5 mm at Level A (P=0.023). Type II endoleak was more frequent in the DS group postoperatively (43.6% vs. 23.8%), but this difference diminished over time. Long-term endoleak rates (>5 years) were significantly higher in the non-DS group (31.4% vs. 7.1%, P=0.009) (Figure S1B). The incidence of dSINE increased over time, with a significantly higher rate in the non-DS group after 5 years (42.9% vs. 10.7%, P=0.002) (Figure S1C).

Patency of FL

FL thrombosis dynamics were evaluated across different anatomical levels (stent-graft, distal DTA, visceral arteries, abdominal aorta, and iliac arteries) over time (Table S2). At 1-year, complete thrombosis was higher at the stent-graft level in the DS group (55.4% vs. 34.8%, P=0.157), and by 5 years, complete thrombosis was 73.2% in DS vs. 47.8% in non-DS (P=0.057), showing a stronger remodeling effect in DS patients (Figure S2A). At 1-year, complete thrombosis at the level of distal DTA reached 8.7% and 3.6% in non-DS and DS groups, while partial thrombosis was higher in the DS group (82.1% vs. 52.2%, P=0.023). This difference persisted at 5 years (78.6% partial thrombosis in DS vs. 65.2% in non-DS, P=0.33) (Figure S2B). By 1 year at visceral arteries level, complete thrombosis remained low in both groups (2.6% vs. 4.3%), while partial thrombosis increased to 39.1% in non-DS vs. 33.9% in DS (P=0.703), showing a slower thrombosis effect in visceral arteries (Figure S2C). At 1 year at level of abdominal aorta, complete thrombosis remained rare (1.8% vs. 0%), while partial thrombosis increased to 25% in DS vs. 26.1% in non-DS (P=0.811) (Figure S2D).

dSINE

Freedom from dSINE was 89.5% (95% CI: 81.9–97.9%) vs. 91.4% (95% CI: 85.2–98%) and 76.9% (95% CI: 65.5–90.3%) vs. 89.4% (95% CI: 82.4–97.1%) at 3 and 5 years, in non-DS and DS groups (P=0.38), respectively (Figure 2E). In the matched cohort, freedom from dSINE was 89.5% (95% CI: 81.9–97.8%) vs. 92.6% (95% CI: 84.5–100%) and 76.9% (95% CI: 65.5–90.3%) vs. 92.6% (95% CI: 84.5–100%) at 3 and 5 years, in non-DS and DS groups (P=0.78), respectively (Figure 2F).

Cumulative incidence of new events—including distal aortic reoperations and dSINE—was significantly lower in the DS group versus non-DS group both before matching (12.9% vs. 29.6%, P<0.001) and after matching (15.5% vs. 29.6%, P=0.04) (Table 3).

Predictors of dSINE and high-risk patients

Univariable Cox-regression analysis identified the risk factors for dSINE as TAA dilatation >45 mm (HR =3.36, 95% CI: 1.32–8.56, P=0.011), Stanford type B AD (HR =3.73, 95% CI: 1.36–10.18, P=0.01), and chronic AD (HR =12.95, 95% CI: 1.52–109.9, P=0.019). Based on multivariable Cox-regression data, the key predictors for dSINE included CTD (HR =3.22, 95% CI: 1.09–9.52, P=0.034), Stanford type B AD (HR =4.3, 95% CI: 1.27–14.61, P=0.019), and chronic AD (HR =7.8, 95% CI: 0.72–84.21, P=0.09). Other factors, including previous aortic repair, type of SG, SG length and size, etc., did not show strong statistical significance (P>0.05) (Figure 3). The Kaplan-Meier analysis demonstrated that patients from DS group had lower risks of dSINE even among high-risk subgroups (Figure S3). While freedom from dSINE was significantly lower in high-risk patients such as those with chronic AD, type B AD, TAA dilation >45 mm, and CTD, the SET showed a protective effect, reducing the incidence of dSINE across all categories. The highest dSINE risk was observed in non-DS group with chronic AD (P=0.009) (Figure S3B), type B AD (P=0.012) (Figure S3D), and aortic dilation >45 mm (P=0.004) (Figure S3F). Additionally, non-DS group with CTD had significantly higher dSINE rates compared to DS group with CTD (P=0.005) (Figure S3H), reinforcing the protective role of SET in this subgroup.

Figure 3 Predictors of dSINE. CI, confidence interval; CTD, connective tissue disorder; DS, dissection-specific; dSINE, distal stent-graft induced new entry; FET, frozen elephant trunk; HR, hazard ratio; PSM, propensity score matching; SG, stent graft; TAAA, thoracoabdominal aortic aneurysm.

Discussion

The SET graft features a tailored radial force distribution, allowing it to adapt to the weakened and fragile intima without exerting excessive pressure that could lead to new tears. The soft distal end, devoid of nitinol rings and featuring a gradual tapering in stiffness, enables deep prosthesis placement, ensuring optimal adaptation to the TL. In our study, we observed significantly lower rates of dSINE compared to conventional FET prostheses in high-risk patients. Although we did not find significant differences in long-term outcomes in the general and PSM cohorts, our subgroup analysis showed a significantly lower incidence of dSINE for SET in high-risk groups. Significantly higher dSINE risk was observed in non-DS group with chronic AD, type B AD, CTD, and TAA dilation >45 mm in comparison with DS group.

dSINE is a frequent complication of FET surgery, with an incidence ranging from 3.4% to 27% in the literature (4-7). The primary risk factors include CTD, chronic AD, and a mismatch between the stent graft size and the TL diameter (8). In most cases, dSINE occurs in the late postoperative period (11 to 36 months post-intervention) and is often asymptomatic, with mortality rates reaching up to 28.6% (9). The mortality rate for treating these complications remains high, reported at 14% in the literature (4). A study by Hiraoka et al. (10) reported a 23% incidence of dSINE after FET procedures for AD, identifying chronic AD and residual FL area as significant risk factors. A review by Jubouri et al. examined the incidence of dSINE and aortic remodeling associated with different FET devices. The study highlighted that Thoraflex Hybrid yielded the lowest incidence of dSINE while promoting favorable aortic remodeling through maximum FL thrombosis and TL expansion. The authors emphasized the importance of FET insertion length and graft size in optimizing outcomes and reducing complications (5). A study by Kreibich et al. evaluated the incidence and risk factors of dSINE after FET procedures. Among 126 patients treated with the Thoraflex hybrid graft, 13% developed dSINE, with a notable increase in risk over time—25% at 36 months. The study found no significant patient-related risk factors but suggested that the stiffer distal ring of the Thoraflex graft may contribute to dSINE. Additionally, patients treated with E-vita had a significantly lower reintervention rate than those with Thoraflex (0% vs. 22%, P=0.003), further supporting the role of graft design in dSINE occurrence (11). Additionally, a systematic review by Nakhaei et al. (12) assessed aortic remodeling, dSINE, and endoleak following FET procedures. The analysis, covering multiple studies, found a mean dSINE incidence of 15.2% post-FET, with the highest risk observed in patients with chronic AD and oversized stent grafts. The study emphasized that device-specific characteristics significantly impact remodeling outcomes, with stiffer grafts increasing dSINE rates. Endoleak occurrence was also linked to inadequate proximal sealing, reinforcing the importance of precise sizing and patient-specific graft selection. Our findings reinforce these observations, demonstrating that the SET prosthesis’s enhanced flexibility and adaptability to the dissected aorta contribute to superior distal remodeling. These results align with recent studies advocating for graft modifications to minimize complications associated with FET.

Study limitations

This study is limited by its retrospective, single-center design. Despite propensity-score matching, some residual bias may remain; however, the density (Figure S4A) and scatter plots (Figure S4B) show good balance. Additionally, since the SET graft was introduced in 2019, follow-up data is relatively short, limiting the ability to assess long-term durability comprehensively. Larger, multicenter studies with extended follow-up are needed to validate these findings.

Future directions

Future research should concentrate on conducting multicenter studies to evaluate the long-term durability of the SET graft. A randomized controlled trial comparing the SET graft with conventional hybrid grafts would provide additional validation of its advantages.

Conclusions

The SET graft represents a substantial advancement in hybrid aortic surgery. Although SET did not emerge as an independent predictor of dSINE in our multivariable analysis, the data demonstrates a protective trend—especially in high-risk subgroups—by addressing the anatomical and biomechanical challenges inherent to AD. Compared to conventional FET devices, the SET graft exhibits superior flexibility, conformability, and an optimized radial-force profile, collectively facilitating enhanced aortic remodeling and reducing the incidence of distal complications. While the overall long-term outcomes between SET and conventional hybrid grafts were comparable in the general and PSM cohorts, subgroup analysis revealed a significant reduction in dSINE rates among patients with chronic AD, CTD, and TAA dilation. These findings support the continued adoption of DS prosthesis as a means to enhance patient outcomes and reduce reintervention rates.

Acknowledgments

None.

Footnote

Funding: None.

Conflicts of Interest: The authors have no conflicts of interest to declare.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Yan TD, Bannon PG, Bavaria J, et al. Consensus on hypothermia in aortic arch surgery. Ann Cardiothorac Surg 2013;2:163-8. [Crossref] [PubMed]
2. Charchyan E, Breshenkov D, Belov Y. A new dissection-specific hybrid stent graft for patients with DeBakey type I aortic dissection. Eur J Cardiothorac Surg 2022;61:596-604. [Crossref] [PubMed]
3. Charchyan ER, Breshenkov DG, Belov YV. The Frozen Elephant Trunk Procedure Using New Russian Hybrid Prosthesis "Soft Elephant Trunk": the Early Experience. Kardiologiia 2023;63:29-37. [Crossref] [PubMed]
4. Kreibich M, Berger T, Rylski B, et al. Aortic reinterventions after the frozen elephant trunk procedure. J Thorac Cardiovasc Surg 2020;159:392-399.e1. [Crossref] [PubMed]
5. Jubouri M, Kayali F, Saha P, et al. Incidence of Distal Stent Graft Induced New Entry vs. Aortic Remodeling Associated With Frozen Elephant Trunk. Front Cardiovasc Med 2022;9:875078. [Crossref] [PubMed]
6. Murana G, Costantino A, Campanini F, et al. Distal stent graft-induced new entry (dSINE) after frozen elephant trunk: a scoping review. Cardiovasc Diagn Ther 2023;13:408-17. [Crossref] [PubMed]
7. Kozlov B, Panfilov D, Lukinov V. Frozen Elephant Trunk for Aortic Dissection Using Different Hybrid Grafts: Preliminary Results from a Prospective Study. J Pers Med 2023;13:784. [Crossref] [PubMed]
8. Li Q, Wang LF, Ma WG, et al. Risk factors for distal stent graft-induced new entry following endovascular repair of type B aortic dissection. J Thorac Dis 2015;7:1907-16. [Crossref] [PubMed]
9. Czerny M, Eggebrecht H, Rousseau H, et al. Distal Stent Graft-Induced New Entry After TEVAR or FET: Insights Into a New Disease From EuREC. Ann Thorac Surg 2020;110:1494-500. [Crossref] [PubMed]
10. Hiraoka T, Furukawa T, Imai K, et al. Distal Stent Graft-Induced New Entry After Frozen Elephant Trunk Procedure for Aortic Dissection. Ann Vasc Surg 2023;97:340-50. [Crossref] [PubMed]
11. Kreibich M, Bünte D, Berger T, et al. Distal Stent Graft-Induced New Entries After the Frozen Elephant Trunk Procedure. Ann Thorac Surg 2020;110:1271-9. [Crossref] [PubMed]
12. Nakhaei P, Bashir M, Jubouri M, et al. Aortic remodeling, distal stent-graft induced new entry and endoleak following frozen elephant trunk: A systematic review and meta-analysis. J Card Surg 2022;37:3848-62. [Crossref] [PubMed]