In 1994, Duke et al. first implemented thoracic aortic endovascular repair (TEVAR), which has since been widely used in clinical practice; its safety and efficacy has also been widely recognized (1, 2). To fully exclude aortic lesions, it is generally recommended that the proximal sealing zone be at least 2 cm, which leads to the need to cover the left subclavian artery (LSA) in some patients (3). However, numerous studies have shown that LSA coverage was associated with the risk of stroke, spinal cord ischemia (SCI), and upper limb ischemia, suggesting the need for revascularization of the LSA during TEVAR (4, 5, 6).

The Society for Vascular Surgery (SVS) 2009 Practice Guidelines recommend routine preoperative LSA revascularization during TEVAR where the proximal seal necessitates coverage of the LSA (GRADE 2, level C). In patients where their anatomy affected critical organ perfusion, routine preoperative LSA revascularization was strongly recommended despite the very low quality of evidence (GRADE 1, level C) (7). However, the 2020 SVS Practice Guidelines state that preoperative or concurrent LSA revascularization is necessary for elective TEVAR (GRADE 1, level B) (8). Traditionally, LSA revascularization was achieved through surgical revascularization of the left carotid artery to LSA bypass or LSA to left carotid artery transposition. Protack et al. included 282 patients undergoing TEVAR and surgical LSA revascularization in a single center retrospective cohort study; the results of the 16-year study showed a 98% primary patency rate at 5 years, low morbidity, and the infrequent need for re-intervention (4). Currently, commonly used endovascular LSA revascularization techniques include the parallel stent technique (chimney, sandwich, and periscope technique), in vitro fenestration technique, in situ fenestration technique, and branched arch endografts (9). Thanks to the continuous research of enterprises and medical centers all over the world, a number of new endovascular devices suitable for a variety of technologies have been entered into clinical research, resulting in rapid progress and promising breakthroughs.

At present, there are no large sample randomized controlled studies or clinical guidelines to indicate which LSA revascularization technique is superior. Therefore, the purpose of this systematic review and meta-analysis was to incorporate the latest clinical studies and compare the safety and efficacy of surgical LSA revascularization and endovascular LSA revascularization during TEVAR.

This analysis was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and was registered with the PROSPERO database on April 16, 2023 (CRD42023414579) (10).

In recent years, several studies have shown that revascularization of the LSA during TEVAR could reduce complications, such as stroke and SCI (27, 28, 29). With the development of new medical devices, more medical institutions have chosen endovascular therapy to reconstruct the LSA. This systematic review and meta-analysis included 14 retrospective cohort studies (1,695 patients) and compared the difference between surgical and endovascular LSA revascularization. The pooled results showed that both treatment methods were safe and effective; there were no significant differences in the outcomes of peri-operative stroke, SCI, type I endoleak, mortality, technical success, and left upper limb ischemia.

A meta-analysis published by Hajibandeh et al. in 2016, which included five studies with a total of 1,161 patients, found that LSA revascularization did not reduce the neurological complications or mortality of patients with LSA coverage (30). However, a relatively small number of studies were included and some relevant studies were omitted. Therefore, the results of the analysis do not fully reflect real world data. Huang et al. added 11 studies in the updated meta-analysis, which included a total of 16 cohort studies (2,591 patients) (31). The pooled results showed that patients who underwent LSA revascularization had a significantly lower incidence of peri-operative stroke than patients who did not undergo LSA revascularization (5.4% vs. 7.8%, P = 0.001). All studies included in the present systematic review and meta-analysis were published after 2016, with peri-operative stroke rates of 3.8% in the surgical LSA revascularization group and 2.6% in the endovascular LSA revascularization group, respectively. Compared with previous studies, the overall incidence rate is still low.

Left carotid-subclavian bypass (CSB) and subclavian transposition (SCT) are traditional methods used for surgical LSA revascularization. Protack et al., from Cleveland Clinic, reported on 282 TEVAR patients who underwent 288 surgical LSA revascularization procedures, with a total of 269 (93%) CSB and 19 (7%) SCT. The incidence of 30-day stroke was 3.5%; the 1, 2, and 5-year primary patency rates were 99.5%, 98.9%, and 98.0%, respectively; and the 1, 5, and 10-year overall survival rates were 82%, 60%, and 42%, respectively. This large sample study conducted over 16 years fully demonstrated the safety and effectiveness of surgical LSA revascularization (4).

Both CSB and SCT techniques provide good patency rates; however, they were both associated with surgical and neurological complications (32, 33). A study by Konstantinou et al. included 211 patients who underwent surgical reconstruction of the LSA either before or after TEVAR. There were 24 cases (11.4%) of local bleeding and 21 cases (10.4%) of re-intervention. Local peripheral nerve injury occurred in 19 patients (9.5%), chylous fistula occurred in eight patients (3.8%), local wound infection occurred in five patients (2.4%), and one patient (0.5%) developed a bypass graft infection. The incidence of stroke was 4.3% (n = 9/211). The incidence of peri-operative stroke in surgical LSA revascularization in the present study was 4.0%, and the incidence of other peri-operative complications was also low. Although there was a higher trend compared to endovascular LSA revascularization, the difference was not significant.

Currently, endovascular LSA revascularization mainly includes parallel stent (chimney and periscope) techniques, fenestration techniques, and single-branched stent techniques. The parallel stent technique was first applied by Crisdo et al. in 2002 as an emergency rescue operation to save LSA blood flow during TEVAR (34). The main advantage of parallel stent technology is that it is not required to be pre-customized and can be used for immediate surgical treatment using commercial stents, resulting in it being commonly used in emergency surgery. The main problem currently affecting the widespread application of the parallel stent technique is the presence of gutter between aortic covered stents, parallel stents, and the aortic wall, leading to type I endoleak. A meta-analysis published by Ahmad et al., which included 11 studies on the use of the chimney graft technique, showed that the incidence of type Ia endoleak ranged from 0 to 20%, and the overall estimated proportion of early type Ia endoleak was 9.4% (95% CI: 6.5%, 13%) (35). Ding et al. reported the results of LSA preservation in patients with type B aortic dissection using the chimney technique; 141 patients underwent LSA revascularization using the chimney technique, and 30 patients (19%) experienced immediate type Ia endoleak (36). In the present study, the incidence of type I endoleak in the endovascular LSA revascularization group was 23.2%. Although there was no significant difference compared to the surgical LSA revascularization group, it remains a problem that deserved attention. However, the subgroup analysis of this study showed that the rate of type I endoleak in the parallel stent group was significantly higher than that in the surgical LSA revascularization group, which may limit the use of parallel stents in the LSA revascularization. In the included literatures, 20 patients in Zhao et al. ’s study received chimney stent implants (14). The 20 patients were divided into a type I endoleak group (n = 11) and a non-type I endoleak group (n = 9). The risk factors for a type I endoleak after chimney stent implantation showed that branch angulation and oversizing were important potential risk factors for postoperative type I endoleak in chimney TEVAR. Most type I endoleak occurred with large branch angles (LSA, >38.8°). In addition, the risk of an endoleak was significantly lower when oversizing of the thoracic stent graft was >11%.

Fenestration techniques include in situ fenestration and pre-fenestration. Mc Williams first reported the success of in situ fenestration in 2004 (37). In 2020, a study by Zhao et al., in China, included 130 patients who underwent in situ laser stent graft fenestration of the LSA during TEVAR. The 5-year follow-up results showed that the LSA patency rate was 97%, and four cases with type I endoleak disappeared during follow-up. There were no neurological complications and no deaths. These results suggest that in situ laser fenestration for LSA revascularization is efficient, safe, and feasible (38). Pre-fenestration included in vitro fenestration and custom fenestration. In 2012, a Japanese study reported the early results of a multicenter clinical trial of 383 patients with a custom fenestration stent (Najuta) for endovascular revascularization of the superior branch of the aortic arch. The technical success rate was 99.22% (n = 380) and the 30-day mortality rate was 1.6% (n = 3). Seven patients experienced cerebrovascular accident (1.8%) and permanent paralysis occurred in three patients (0.8%). Furthermore, ascending aorta dissection was observed in three patients (0.8%) (39). Fenestration with TEVAR was performed with a high success rate and low complication rate for the patients with LSA revascularization. However, the position and size design of the fenestrated stent and the precise release of the stent during the operation requires strict control by the surgeon. Furthermore, the structure of the stent was damaged by fenestration, the long-term safety needs to be further studied.

The single-branch stent technique was first proposed by Inoue et al. in 1996, and was successfully used to repair Stanford type B aortic dissection involving the LSA. Subsequently, they reported the early and mid-term results of 17 patients treated with the Inoue single-branched stent for thoracic aortic lesions involving the LSA in 2005, demonstrating the effectiveness of the Inoue single-branched stent (40, 41). A meta-analysis published in 2022 investigated the use of the Castor stent for LSA revascularization (42). The study included 11 articles with a total of 415 patients, and the pooled results showed a technical success rate of 97.5%. The early type I endoleak rate was 1.6%, the 30-day mortality rate was 0.96%, the early re-intervention rate was 0.9%, the incidence of perioperative stroke was 0%, and the 1-year survival rate was 99.7%. The results of this study showed that use of the Castor stent to reconstruct the LSA during TEVAR was feasible and effective.

The main limitations of this systematic review and meta-analysis include the following points. Firstly, the included studies were all retrospective studies; therefore, this study is lacking in RCTs. Secondly, there was a deficiency in long-term follow-up data. Thirdly, the number of articles included in this study was too small to conduct subgroup analyses. Finally, one of the reasons for the absence of a statistically significant difference between these two treatment arms could be the low number of patients in the endovascular LSA revascularization group.

There was no significant difference in terms of short-term outcomes when comparing the surgical and endovascular LSA revascularization techniques. The quality of evidence assessed by GRADE scale was low to very-low. Use of both techniques during TEVAR was found to be safe and effective. Compared with surgical LSA revascularization techniques, parallel stent revascularization of the LSA significantly increased the rate of type 1 endoleak.