Mid-term efficacy of castor stent and in situ fenestration stent in the treatment of type B aortic dissection involving the left subclavian artery: A retrospective single-center study

Abstract

To evaluate the midterm efficacy of the Castor stent (CS) versus in situ fenestration (ISF) for reconstructing the left subclavian artery (LSA) in patients with type B aortic dissection (TBAD).

Between July 2017 and July 2022, a total of 247 patients with TBAD were enrolled. One hundred thirty-seven patients were treated using CSs (group A), while the remaining 110 patients received ISFs (group B). Data of the two groups were retrospectively analyzed.

The success rates of surgery were 99.3% and 95.5% in groups A and B (p = .053), There were no deaths during hospitalization. During surgery, group B showed a longer surgical duration [68.0 (66.0, 77.0) vs. 62.0 (59.0, 66.0) min, p < .001] and intraoperative fluoroscopy time [18.0 (16.0, 20.0) vs. 16.0 (14.0, 18.0) min, p < .001] than group A. The follow-up duration was similar for both groups (44.0 vs. 43.0 months, p = .877), and no patient died. Stent-related complications were significantly lower in group A than in group B (1.5% vs. 8.4%, p = .009). Group A had fewer instances entry flow (0.7% vs. 4.7%, p = .048) and stent stenosis (0.7% vs. 2.8%, p = .206) than group B. All reintervention cases (4.7%) were from group B (p = .011). The rate of false aortic lumen thrombosis was significantly higher in group A than in group B (84.6% vs. 72.9%, p = .024).

Both CSs and ISFs are evidently safe, feasible, and effective in achieving positive early outcomes in patients undergoing treatment for TBAD. Notably, at midterm follow-up, CSs appeared to be superior to ISF in terms of reducing stent-related complications and minimizing the need for reintervention.

INTRODUCTION

With advances in thoracic endovascular aortic repair (TEVAR) for treating aortic dissection (AD), TVAR has become the standard treatment for managing both acute and chronic type B aortic dissection (TBAD).^1–3 For TBAD, wherein the landing zone of the aortic arch is insufficient or involves the left subclavian artery (LSA), LSA reconstruction is currently the first choice in most centers and has been proven safe, effective, and superior to direct closure of the LSA.^4–7 At our center, the most common surgical methods employed include the placement of Castor stents (CSs) and in situ fenestration (ISFs).^8,9 These techniques not only maintain the blood supply to the LSA covered by the aortic stent and provide additional landing sites for the main stent graft.¹⁰ Accumulating evidence indicates that most patients treated using these two methods have experienced favorable short-term outcomes.^11–14 However, few studies have compared the efficacy, complication rates, and prognosis of these two procedures at midterm follow-up. Therefore, this study aimed to compare the midterm efficacy of CSs and ISFs for TBAD, particularly when lesions involved the LSA, or the landing zone of the aortic arch was insufficient.

METHODS Study population

Consecutive patients diagnosed with TBAD who underwent TEVAR at our hospital between July 2017 and July 2022 were retrospectively enrolled. We included patients aged 18−80 years who underwent TEVAR under local or general anesthesia and subsequently underwent LSA reconstruction through placement of either CSs or the use of the ISF technique. The exclusion criteria were as follows: (i) patients with aortic dissection involving the ascending aorta, innominate artery, or left common carotid artery; (ii) those who had simultaneous reconstruction of the brachiocephalic artery or left common carotid artery during surgery; (iii) individuals diagnosed with Marfan syndrome; (iv) patients with coagulation disorders; (v) those with chronic renal failure, acute or chronic infectious diseases, or systemic immune diseases undergoing anti-immunosuppressive or anti-inflammatory therapy; and (vi) individuals who underwent additional surgeries simultaneously.

Clinical data

The data collected from the electronic medical record system of our hospital included baseline information, medical histories, operative details, postoperative outcomes, and follow-up details. Patient follow-ups were conducted through in-person outpatient visits, follow-up center appointments, telephone consultations, WeChat communications, or message-based communications. All patients underwent regular thoracic aortic computerized tomography angiography at 3 months, 1 year, and annually after discharge. Additionally, all patients were administered antiplatelet therapy (enteric-coated aspirin, 100 mg/day) for the initial 3 months immediately after TEVAR

Definitions

Multiple organ ischemia was defined as ischemia of two or more vital organs. Stent-related complications were defined as issues associated with stents implantation, including entry flow, stent migration, stent stenosis, stent occlusion, and mediastinal infection. Surgical success was defined as commencing with the establishment of a vascular approach on an intent-to-treat basis leading to the successful introduction and deployment of stents, without converting to an open repair operation. Furthermore, there was no entry flow, stent stenosis, migration or occlusion, and no death within 24 h of surgery. False aortic lumen thrombosis was characterized by the absence of blood flow in the false aortic lumen of the stent segment, as revealed by computed tomography angiography imaging. Type I entry flow was defined as flow between the aortic wall and the proximal stent flow, either into the proximal false aortic lumen and primary entry tear (Type IA) or flow into the distal entry tear and distal false aortic lumen (Type IB). Type II entry flow was defined as retrograde entry flow through thoracic bronchial or arch vessel branches (innominate artery, left carotid artery, left subclavian artery) or intercostal arteries into the false aortic lumen. Type R entry flow was specified as true lumen flow with antegrade entry into the false aortic lumen through distal uncovered branch fenestrations, including intercostal, visceral, renal, lumbar arteries and iliac branch arteries, or septal fenestrations, excluding type IB¹⁵ The follow-up data included information related to death, cardiac events, cerebral events, stent-related complications, false aortic lumen thrombosis, and reintervention.

Operational procedure

Following either local anesthesia with lidocaine or general anesthesia, femoral artery access was established using a vascular suture device (Perclose ProGlide system, Abbott Vascular, CA, USA) or through a femoral artery incision. A 1 cm incision was then made to expose the body surface of the left brachial artery to expose it, and a 5F vascular sheath was inserted for additional support. Subsequently, the entire body received heparinization at a dose of 1 mg/kg. Using the Seldinger technique, a “pigtail” catheter was inserted through the femoral artery to the ascending aorta for angiography to identify the location and involvement of the dissection break. The super stiff guide wire (COOK, Bloomington, IN, USA) and vascular sheath were exchanged, carefully threading the super stiff wire to reach the ascending aorta.

For the ISFs technique, we used an Ankura stent graft (Lifetech Scientific Company, Ltd., Shenzhen, China) guided by the superstiff wire to reach the aortic arch. The proximal end of the stent was deployed at the distal end of the left common carotid artery opening and released under fluoroscopy to cover the LSA (Figure 1A). A Fustar steerable vascular sheath (Lifetech Scientific Co., Ltd., Shenzhen, China) was inserted through the left brachial artery, and a balloon-expandable puncture needle (Lifetech Scientific Co., Ltd., Shenzhen, China) was introduced through the Fustar steerable vascular sheath. The puncture direction and position were determined by rotating the C-arm radiator and aligning it with the “8” and “O” shapes on the Ankura stent graft. Subsequently, the membrane of the Ankura stent was punctured using a balloon puncture needle, and a guidewire was carefully advanced to the descending aorta through the stent membrane puncture site (Figure 1B). The Fustar steerable vascular sheath was then introduced into the Ankura stent lumen along the guidewire. Based on the size of the individual LSA, balloon catheters of different sizes were used to expand the number of puncture points from small to large (Figure 1C). Through the guide wire, the Fluency stent graft (C.R. Bard, Inc., Murray Hill, NJ, USA) or Viabahn stent graft (Gore & Associates, Flagstaff, AZ, USA) was deployed through the fenestration into the Ankura stent lumen and the LSA. The proximal end extended into the Ankura stent lumen by 5−10 mm (Figure 1D). The selected size was larger than the original vessel diameter, the diameter of the left subclavicular artery lumen was generally selected to be 10%−20% larger,¹⁰ and the stent was released under fluoroscopy once the positioning was confirmed. Finally, the stent was further expanded using a balloon to ensure a precise fit within the Ankura stent graft.

FIGURE 1. The operational procedure of in situ fenestration stent implantation.

For CSs implantation, after completing aortography, a Castor stent graft (MicroPort Medical, Shanghai, China) was advanced to the descending aorta through the femoral artery sheath along the super-stiff wire (Figure 2A). The branch stent was directed into the LSA by maneuvering the lead wire into the vascular sheath of the left brachial artery (Figure 2B). Subsequently, a guidewire was used to pull it out of the vascular sheath of the left brachial artery. After the trigger line was drawn under fluoroscopy, the principal part of the CSs was swiftly deployed, and the LSA branch stent was deployed by extracting the traction line (Figure 2C).

FIGURE 2. The operational procedure of Castor stent implantation.

Angiography was performed immediately after stent implantation in both groups to confirm unobstructed blood flow in the stent, proper positioning of the stent, absence of entry flow, and complete exclusion of aortic dissection (Figures 1D and 2D). Subsequently, the catheter, guidewire, and vascular sheath were carefully removed, and hemostasis measures were meticulously carried out. The femoral artery and left brachial artery were sutured, and no bleeding was observed. The skin and fascia were then sutured to these layers, marking the conclusion of the surgery.

Statistical analysis

Categorical variables are presented as numbers or percentages, while continuous variables are presented as medians (Q25, Q75) or means ± standard deviations. The two groups were compared using the U test or t test for continuous variables and the chi-squared test for categorical variables. SPSS 22.0 (IBM^®, Armonk, NY, USA) was used for data analysis. Significance was set at p < .05.

RESULTS

During the study period, CSs were placed in 137 patients (group A), and ISFs was implemented in 110 patients (group B). Both groups had similar preoperative demographic characteristics (Table 1).

TABLE 1 Patient characteristics of the two groups.

Characters	Group A (n = 137)	Group B (n = 110)	p value
Age (years)	52.0 (45.0, 61.0)	54.5 (48.0, 62.0)	.195
Male sex, n (%)	89 (65.0)	69 (62.7)	.818
Hypertension, n (%)	101 (73.7)	84 (76.4)	.743
Diabetes mellitus, n (%)	4 (2.9)	6 (5.5)	.315
Coronary artery disease, n (%)	6 (4.4)	4 (3.6)	.768
COPD, n (%)	5 (2.6)	3 (2.7)	.684
Renal insufficiency, n (%)	2 (1.5)	0 (0.0)	.203
Stroke history, n (%)	2 (1.5)	2 (1.8)	.825
Previous aortic surgery, n (%)	3 (2.2)	4 (3.6)	.496
Visceral ischemia, n (%)	0 (0.0)	0 (0.0)	N/A
Limb ischemia, n (%)	0 (0.0)	0 (0.0)	N/A
Renal ischemia, n (%)	2 (1.5)	2 (1.8)	.825
Spinal cord ischemia, n (%)	0 (0.0)	0 (0.0)	N/A
Multiple organ ischemia, n (%)	0 (0.0)	0 (0.0)	N/A
Phase of TBAD
Acute, n (%)	19 (13.9)	14 (12.7)	.941
Subacute, n (%)	98 (71.5)	80 (72.7)	.948
Chronic, n (%)	20 (14.6)	16 (14.5)	.865
Dissection of LSA, n (%)	74 (54.0)	62 (56.4)	.810

Continuous variables were present as medians (Q25,Q75). U or t test for continuous variables and chi-square test for categorical variables.

Abbreviations: COPD, chronic obstructive pulmonary disease; LSA, left subclavian artery; TBAD, type B aortic dissection.

Operative data

The operative data of the two groups are summarized in Table 2. The surgical success rates in groups A and B were 99.3% and 95.5% (p = .053), respectively, and there were no reported deaths during the hospitalization. One patient in group A was switched to the chimney technique because the LSA originated from the lateral posterior wall of the aortic arch and the angle with the aortic arch was too small, which could impair the accurate release of the principal part of the CS after rotation. Conversely, in group B, the ISF procedure was discontinued in three patients, and the treatment team chose to occlude the LSA because the Ankura stent membrane could not be accurately located and successfully punctured. In two patients, the membrane was torn at the edge of the metal support stent due to balloon expansion after multiple punctures, resulting in a “chimney” shape after stent implantation. The operative duration was longer in group B than in group A [68.0 (66.0, 77.0) vs. 62.0 (59.0, 66.0), p < .001], and the intraoperative fluoroscopy time was also longer in group B than in group A [18.0 (16.0, 20.0) vs. 16.0 (14.0, 18.0), p < .001].

TABLE 2 Intraoperative and postoperative findings in the two patient groups.

Valuables	Group A (n = 137)	Group B (n = 110)	p value
Intraoperative data
Local anesthesia, n (%)	16 (11.7)	9 (8.2)	.488
General anesthesia, n (%)	121 (88.3)	101 (91.8)	.488
Femoral artery incision, n (%)	49 (35.8)	31 (28.2)	.259
Vascular suture device, n (%)	88 (64.2)	79 (71.8)	.259
Viabahn branch, n (%)	0 (0.0)	34 (30.9)	N/A
Fluency branch, n (%)	0 (0.0)	76 (69.1)	N/A
Operative time (minutes)	62.0 (59.0, 66.0)	68.0 (66.0, 77.0)	<.001
Fluoroscopy time (minutes)	16.0 (14.0, 18.0)	18.0 (16.0, 20.0)	<.001
Occlusion of LSA, n (%)	0 (0.0)	3 (2.7)	.052
Chimney procedure, n (%)	1 (0.7)	2 (1.8)	.438
Surgical success, n (%)	136 (99.3)	105 (95.5)	.053
Postoperation data
Early death, n (%)	0 (0.0)	0 (0.0)	N/A
Fever, n (%)	25 (18.2)	23 (20.9)	.716
Stroke, n (%)	0 (0.0)	0 (0.0)	N/N
Paraplegia, n (%)	0 (0.0)	0 (0.0)	N/A
Acute renal failure, n (%)	1 (0.7)	0 (0.0)	.369
Early entry flow, n (%)	0 (0.0)	0 (0.0)	N/A
ICU stay (hours)	13.0 (6.0, 16.0)	13.0 (8.0, 17.0)	.523
Length of stay (days)	6.0 (5.0, 6.0)	6.0 (5.0, 7.0)	.162

Continuous variables were present as median (Q25,Q75) or mean ± SD. Chi-square test for categorical variables and t test or Wilcoxon rank sum test for continuous variables.

Abbreviations: ICU, intensive care unit; LSA, left subclavian artery.

Postoperative data

All patients were safely transferred to the intensive care unit or inpatient ward for postoperative rehabilitation. Notably, there were no new occurrences of acute renal failure, stroke, paraplegia, early entry flow, or death in either group during the early postoperative period. The lengths of stay in the intensive care unit and the hospital were comparable between the two groups [13.0 (6.0, 16.0) vs. 13.0 (8.0, 17.0) days, p = .523] hours and [6.0 (5.0, 6.0) vs. 6.0 (5.0, 7.0) days, p = .162] days, respectively). Importantly, all patients recovered and were subsequently discharged from the hospital (Table 2).

Follow-up data

Three patients (2.7%) in group B were lost to follow-up. Both groups had similar postoperative follow-up durations [44.0 (28.5, 56.0) vs. 43.0 (26.0, 58.0), p = .877]. During the follow-up period, no deaths were reported in either group. One case of (0.7%) entry flow was observed in group A, while in group B, there were 5 cases (4.7%) of entry flow (p = .048), including one case of (0.9%) of type I entry flow, one case (0.9%) of type II entry flow, and three cases (2.7%) of entry flow from disconnection between stents. The incidence of stent stenosis was 0.7% and 2.8% in groups A and B, respectively (p = .206).

Overall, the two groups had significantly different incidences of stent-related complications (group A: 1.5% vs. group B: 8.4%, p = .009). The rate of false aortic lumen thrombosis was higher in group A than in group B (84.6% vs. 76.6%, p = .370). No patients in group A needed reintervention, whereas five patients (4.7%) in group B underwent reintervention (p = .011), including three with entry flow from disconnection between stents, one with mediastinal infection, and one with ascending aorta replacement due to ascending aorta dissection (Table 3).

TABLE 3 Follow-up data of the two patient groups.

Valuables	Group A (n = 137)	Group B (n = 107)	p value
Follow-up time (months)	44.0 (28.5, 56.0)	43.0 (26.0, 58.0)	.877
Follow-up death, n (%)	0 (0.0)	0 (0.0)	N/A
Stent-related complications, n (%)	2 (1.5)	9 (8.4)	.009
Entry flow, n (%)	1 (0.7)	5 (4.7)	.048
Type I entry flow, n (%)	0 (0.0)	1 (0.9)	.257
Type II entry flow, n (%)	0 (0.0)	1 (0.9)	.257
Type R entry flow, n (%)	0 (0.0)	0 (0.9)	N/A
Entry flow from disconnection between stents, n (%)	1 (0.7)	3 (2.8)	.448
Stent stenosis, n (%)	1 (0.7)	3 (2.8)	.206
Stent migration, n (%)	0 (0.0)	0 (0.0)	N/A
Stent occlusion, n (%)	0 (0.0)	0 (0.0)	N/A
Mediastinal infection, n (%)	0 (0.0)	1 (0.9)	.257
Aortic FLT, n (%)	116 (84.6)	85 (76.6)	.370
Chronic renal failure, n (%)	1 (0.7)	0 (0.0)	.376
Stroke, n (%)	0 (0.0)	0 (0.0)	N/A
AAD, n (%)	0 (0.0)	1 (0.9)	.257
Reintervention, n (%)	0 (0.0)	5 (4.7)	.011

Continuous variables were present as medians (Q25,Q75) or means ± SD. Chi-square test for categorical variables and t test or U test for continuous variables.

Abbreviations: AAD, ascending aorta dissection; FLT, false lumen thrombosis.

DISCUSSION

This study is the first to compare the early- and mid-term outcomes between the placement of CSs and the ISF technique in patients with TBAD who exhibited insufficient anchoring in the aortic arch or LSA involvement. Remarkably, the patients recovered rapidly after surgery and benefited from endovascular treatment. Our findings underscore the viability and efficacy of both CSs and ISFs techniques, particularly in the early and middle stages of patient care.

In this study, surgery and fluoroscopy times were shorter in group A than in group B. Several factors contributed to this finding. First, in group A, the main technical challenge was establishing the left brachial artery and femoral artery track. During the surgery, the guidewire was inserted through the vascular sheath of the femoral artery and guided anteriorly through the LSA into the vascular sheath of the left brachial artery.^16,17 A track was created from the left brachial artery to precisely guide the branch of the CSs into the LSA precisely. This process can be performed in a relatively short period using the roadmap mode technique or an arresting device through the sheath of the left brachial artery. In group B, we had to insert a balloon-expandable puncture needle with the assistance of the Fustar steerable vascular sheath. A contrast agent-filled balloon was used to secure the Ankura stent graft, which was then used for in situ fenestration, and the puncture point was gradually expanded through the balloon.^18,19 Finally, a branch stent was implanted. The ISF technique is thus was more complicated than the CS technique, resulting in longer operation and fluoroscopy times. Second, in the CS procedure, the aortic arch was fine-tuned with the CS delivery system, and the stent was adjusted to ensure an optimal fit between the branch angle and the original blood vessel. The difficulty of the operation for CS placement was greatly reduced, and the operation and fluoroscopy times were reduced. In contrast, in the ISF procedure, the Fustar steerable vascular sheath is used in combination with adjusting the direction of the C-arm radiator and the orientation of the “8” and “O” shapes on the Ankura stent to achieve proper positioning.²⁰ Thus, the puncture needle was guided to maintain a median vertical position with the main stent for puncture. However, the angle between the LSA and aortic arch often deviated from the ideal position in most patients. In some patients, the LSA originated from the lateral wall of the aortic arch, significantly complicating the identification of the puncture point and leading to longer operation and fluoroscopy times. Third, in the CSs procedure, when the whole stent was fully opened, the guidewire was extracted, and the marked catheter was advanced to the proximal end of the aorta for angiography. This angiography confirmed critical parameters such as the unobstructed blood flow in the stent, proper position of the stent, absence of entry flow from disconnection between stents, and complete exclusion of aortic dissection. Subsequently, surgery was completed. However, in the ISFs procedure, partial branch stents were not opened satisfactorily, leading to branch stenosis at the main stent graft puncture site. As a result, it was necessary to re-inflate the branched stent using a balloon, prolonging the operative and irradiation times. In summary, compared with ISFs, CSs is a simpler surgery, resulting in shorter operation and fluoroscopy times.

In this study, the success rate was quite high for both groups due to the short learning curve of the technology and the mastery of both methods by our surgeons. However, our findings suggest that the ISFs technique is more selective for patients, whereas the CSs technique is more applicable. This is primarily because the LSA often has different angles in the aortic arch.²¹ In patients with too small or too large angles between the LSA and the aortic arch and those with the LSA originating from the side wall of the aortic arch,²² the difficulty of puncturing the stent membrane is greatly increased, the accuracy of the puncture location is reduced, and the surgical risk is elevated during surgery. Conversely, CSs allow for aortic arch adjustments, and most angles between the LSA and the aortic arch can be accommodated, thereby increasing the applicability of the technique and enhancing surgical success rates. In some patients with a steep aortic arch, the ISFs technique has certain limitations, particularly the insufficient anchorage distance of the main stent graft.²³ Insufficient anchorage distance necessitates moving the main stent graft forward, potentially covering or partially covering the left common carotid artery, which may affect its flow.²⁴ Notably, CSs have a variety of models, and an appropriate stent type can be chosen according to the morphology of the LSA and aortic arch, greatly increasing adaptability and surgical success rates. Additionally, during CS placement, when the stent does not match the original blood vessel or the position is not satisfactory during the operation, we can retract the bracket before it is fully opened. Conversely, when in situ fenestration fails or the angle is unsatisfactory, we cannot retract the stent. Consequently, the LSA opening is occluded, or the chimney technique is performed.²⁵

During the follow-up period, we found that stent stenosis and entry flow from disconnected stents was more common in the ISFs group than in the CSs group. In the CSs approach, the branch and main stent grafts were integrated and well anastomosed. After accurate intraoperative positioning, the CSs could be fully opened and better adapted to the original vascular morphology, significantly reducing the incidence of branch stent stenosis and entry flow from disconnection between stents.¹¹ Conversely, in the ISFs group, the opening of the branch stent was determined not only by the expansion degree of the balloon to the puncture point but also by the puncture position, particularly the position near the metal stent. This proximity restricted the opening of the branch stent due to the presence of the metal structure of the main stent, resulting in neck occlusion and subsequent stent stenosis. Furthermore, when the angle between the LSA and the aortic arch was too large or too small, the angle between the branch stent and the main stent graft would also be too large or too small to adapt to the original vascular angle upon the release of the branch stent. This mismatch contributed to stent stenosis or loose connections, increasing the risk of entry flow from disconnection between stents.

In this study, we also found that the rate of false aortic lumen thrombosis was higher in the CSs group (84.6 %) than in the ISFs group (76.6%), although this difference was not statistically significant. Some studies have suggested that the false aortic lumen may only be completely isolated when the true lumen is completely isolated.^26–29 It is plausible that the slightly higher probability of false aortic lumen thrombosis in the CSs group than in the ISFs group could be attributed to the relatively small number of cases in this study. Further studies involving a larger sample size are needed to comprehensively understand the differences between the two methods in reducing the probability of false aortic lumen thrombosis.

Limitations

This study was a retrospective analysis with a limited sample size, which might limit statistical power. Additionally, the study's follow-up time was limited, underscoring the need for further assessment of the long-term efficacy of these two techniques. Moreover, this study did not observe molding of the residual dissection of the descending and abdominal aorta after stent implantation in either group, which might also have introduced a statistical bias while evaluating the reintervention and false aortic lumen thrombosis rates in this study. Further prospective controlled studies are imperative to comprehensively understand the distinct suitability of different stents for varying angles or shapes of the LSA and aortic arch.

CONCLUSIONS

Both CSs and the ISFs technique were safe, feasible, and effective in achieving favorable early outcomes in patients with TBAD. Notably, CSs could reduce surgical and fluoroscopy times; and at midterm follow-up, CSs appeared to be superior to ISFs in terms of reducing stent-related complications, such as stent stenosis, entry flow, and mediastinal infection, as well as minimizing reintervention events. A comprehensive assessment of their long-term efficacy is needed for a more complete understanding of their benefits and limitations.

AUTHOR CONTRIBUTIONS

Liangwan Chen and Zhihuang Qiu designed the study and submitted the manuscript. Qingsong Wu and Lingfeng Xie prepared the first draft of the manuscript and made the literature review. Qingsong Wu and Lingfeng Xie are contributed equally to this study and share first authorship. Huangwei Li made substantial changes in the manuscript. Yue Shen collected and analyzed data together. All authors read and approved the final manuscript.

ACKNOWLEDGMENTS

We would like to thank Editage (www.editage.cn) for English language editing. This work was funded by the National Natural Science Foundation of China (U2005202), the Natural Science Foundation of Fujian Province (2020J02056), Key Laboratory of Cardio-Thoracic Surgery (Fujian Medical University), Fujian Province University (No.2019-067) and Startup Fund for Scientific Research at Fujian Medical University (2022QH2019).

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no conflict of interest.

DATA AVAILABILITY STATEMENT

All data generated or analyzed during this study are included in this published article.