Comparison of short-and long-term outcomes between endovascular and open repair for descending thoracic aortic aneurysm: a systematic review and meta-analysis

Abstract

Objective:

This systematic review and meta-analysis aimed to evaluate and compare the efficacy of endovascular versus open repair for the treatment of patients with descending thoracic aortic aneurysm (DTAA).

Methods:

A systematic search of the PubMed, Embase, and Cochrane Library databases for relevant studies was performed. Outcome data, including postoperative mortality and morbidity, operative details, all-cause survival, freedom from aortic-related survival and freedom from aortic-related re-intervention, were independently extracted by two authors in a standardized way.

Results:

Twenty-nine studies comprising 49 972 patients (22 049 endovascular repair; 27 923 open repair) were included. Endovascular repair was associated with a significantly lower postoperative mortality rate [odd ratio (OR): 0.57, 95% confidence interval (CI): 0.45-0.72; I² = 72.58%] and morbidity. In terms of long-term survival, endovascular repair yielded better freedom from aortic-related survival [hazard ratio (HR): 0.71, 95% CI: 0.54-0.93, P = 0.012] but inferior freedom from aortic-related reintervention (HR: 2.10, 95% CI: 1.45-3.04, P < 0.001). Landmark analysis revealed that the open repair group experienced better all-cause survival beyond 16 months (HR: 1.64, 95% CI: 1.53-1.75, P < 0.001). In addition, in the subgroup of patients with intact DTAA, those who underwent open repair exhibited a higher rate of postoperative mortality (OR: 0.58, 95% CI: 0.38-0.88; I² = 83.34%) but had better all-cause survival beyond 7 months (HR: 1.72, 95% CI: 1.61-1.84, P < 0.001) than those who underwent endovascular repair.

Conclusion:

Among patients treated for DTAA, endovascular repair was associated with better freedom from aortic-related survival, a lower risk for postoperative mortality and morbidity, and shorter lengths of intervention, intensive care unit stay, and hospital stay than those who underwent open repair. Open repair yielded significantly better long-term all-cause survival and freedom from aortic-related re-intervention than endovascular repair.

Introduction

Open surgical repair is gold-standard treatment for descending thoracic aortic aneurysms (DTAA)^[1]. In 1994, thoracic endovascular aortic repair (TEVAR) was recommended as an effective treatment for DTAA and approved by the United States Food and Drug Administration in 2005^[2,3]. Since then, the application of TEVAR has rapidly increased, revolutionizing the treatment of DTAA, and has been recommended as the first-line option for low-risk patients^[4,5]. Studies have reported that TEVAR reduces procedure time, infection risk, and operative mortality, but also significantly increases the risk for serious complications, including endoleak, stent-graft infection, new-onset dissection, and stent-graft migration^[3,6-9].

However, randomized controlled trials comparing the effects of endovascular and open surgical repair for DTAA remain lacking. Results of existing meta-analyses investigating the efficacy of TEVAR are controversial^[10-12]. In addition, the long-term outcomes of these 2 procedures are unclear. As such, the present meta-analysis aimed to systematically collect relevant studies comparing the short- and long-term outcomes of endovascular versus open surgical repairs for DTAA to provide a more reliable basis for clinical decision making.

Methods

Search strategy

The present analysis has been reported in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (i.e., “PRISMA”) statement and Assessing the methodological quality of systematic reviews (i.e., “AMSTAR”) Guidelines^[13,14], and has been registered on the PROSPERO website. Two reviewers independently searched the PubMed, Embase, and Cochrane Library databases for relevant studies published up to June 2024 (Supplementary Table 1, available at: http://links.lww.com/JS9/D775).

The following search terms were used: “thoracic aortic aneurysm,” “open,” “endovascular,” and “TEVAR.” The search strategy was confined to studies involving human subjects, but not solely to the English language literature.

Eligibility criteria

Studies fulfilling the following criteria were included: enrolled patient population was diagnosed with DTAA; comparative studies in which patient groups were treated with endovascular repair and open repair; and, for studies using the same data source, the most recent or complete studies were included.

Studies that enrolled patients with non-aneurysmal descending thoracic aortic diseases (e.g., aortic dissection, pseudoaneurysm, intramural hematoma), nonhuman studies, reviews, meta-analyses, case reports, and those that did not report the primary outcomes of interest (i.e., postoperative mortality or long-term survival), were excluded.

Data extraction and quality assessment

Data from the included studies were independently extracted by two authors using a standardized protocol. Any disagreements were resolved through consensus discussion or assistance from a third author. The following data were extracted: first author’s name; year of publication; country; sample size; demographic data; clinical characteristics; and short-term outcomes, including hospital stay, length of intensive care unit (ICU) stay, intervention length, postoperative morbidity and mortality; and long-term survival outcomes, including all-cause survival, freedom from aortic-related survival and aortic-related reintervention.

Two authors independently evaluated the quality of included studies based on the Newcastle-Ottawa Scale^[15]. The total quality score of the studies ranged from 0 to 9, and a score >5 was considered to represent high quality. Two authors independently assessed the certainty of evidence according to the Grading of Recommendations Assessment, Development, and Evaluations (i.e., “GRADE”) system^[16]. GRADEpro GTP (www.gradepro.org) was used to generate summary of findings tables.

Statistical analysis

Short-term outcomes are expressed as odds ratio (OR) for dichotomous variables or weighted mean difference (WMD) for continuous variables with corresponding 95% confidence interval (CI). Statistical heterogeneity among the included studies was assessed using the Q and the I-squared (I²) statistics. Studies with P >0.10 or I² <50% were considered to have low heterogeneity^[17], and a random-effects model was used to pool study estimates. Meta-regression analysis was used to explore the sources of heterogeneity. Sensitivity analysis was used to test the robustness of the primary outcomes. Egger’s test was used to evaluate publication bias, with significant bias defined as P <0.05.

Kaplan–Meier (KM) curves reporting all-cause survival, freedom from aortic-related survival, and freedom from aortic-related reintervention among subjects treated with endovascular or open repair were extracted from each study, and estimated individual patient data (IPD) were reconstructed using the IPD from the KM method, as described by Liu et al^[18]. In the first stage, the points in each KM curve were manually selected to extract the data coordinates (time and survival probability). In the second stage, the estimated IPD was reconstructed based on the extracted data coordinates.

Finally, the estimated reconstructed IPD of the time-to-event data from all studies were presented using KM curves and compared using the log-rank test. The hazard ratio (HR) with corresponding 95% CI for the difference between the two treatment groups was calculated using a Cox proportional hazards regression model. The proportionality of the hazards of the Cox model was checked using the Grambsch–Therneau test and diagnostic plots based on Schoenfeld residuals^[19]. Landmark analysis and flexible parametric survival models with B-splines were performed based on either visual or statistical violations of hazard proportionality^[20]. In addition, the difference in the restricted mean survival time (RMST) over time was modeled^[21].

Differences with P <0.05 were considered to be statistically significant. Statistical analyses were performed using Stata/MP Release 17.0 (StataCorp LLC, College Station, TX, USA) and RStudio version 4.2.1 (RStudio Team [2020]. RStudio: Integrated Development for R. RStudio, PBC, Boston, MA, USA; www.rstudio.com).

Results

Study selection

The initial online literature search retrieved 4233 studies and, after eliminating 896 duplicates, 3337 potentially eligible studies were selected for further screening (Fig. 1). After screening titles, abstracts, and full texts, 29 studies fulfilled the predefined inclusion criteria were included in the meta-analysis^[22-50]. All of the included studies were cohort studies.

Figure 1.

PRISMA flow diagram of the process for the identification of eligible studies.

Study characteristics and quality

The included studies were published between 1998 and 2023, and comprised 49 972 patients with DTAA, of whom 22 049 underwent endovascular repair and 27 923 underwent open surgical repair. Among these studies, patient data for 8 were sourced from databases. Characteristics of the included studies and patient populations are summarized in Table 1 and Supplementary Table 2 (available at: http://links.lww.com/JS9/D775).

Table 1.

Characteristics of studies

Study	Year	Country	Data source	Study period	Aneurysm type	Aneurysm location
Ehrlich et al[21]	1998	Austria	Single center	1989-1997	Mixed	Descending
Najibi et al[22]	2002	US	Multi center	1999-2000	Intact	Descending
Glade et al[23]	2005	The Netherlands	Multi center	1997-2003	Intact	Descending
Fairman et al[24]	2008	US	Multi center	2003-2005	Intact	Descending
Makaroun et al[25]	2008	US	Multi center	1999-2001	Intact	Descending
Patel et al[26]	2008	US	Single center	1993-2007	Intact	Descending
Kieffer et al[27]	2009	France	Single center	1997-2005	Intact	Descending
Andrassy et al[29]	2011	Germany	Single center	1992-2008	Mixed	Descending
Arnaoutakis et al[30]	2011	US	Single center	2000-2009	Mixed	Descending
Jonker et al[32]	2011	US	Multi center	1995-2010	Ruptured	Descending
Desai et al[33]	2012	US	Single center	1999-2007	Mixed	Descending
Karimi et al[34]	2012	US	Single center	2005-2007	Intact	Descending
Matsumura et al[35]	2014	US	Multi center	2004-2006	Intact	Descending
Arnaoutakis et al[37]	2015	US	Single center	2002-2013	Intact	Descending
Gillen et al[38]	2015	US	Single center	2005-2012	Intact	Descending
Lee et al[39]	2015	Korea	Single center	2006-2013	Mixed	Descending
Shiraev et al[40]	2016	Australia	Multi center	2003-2013	Mixed	Descending
Tanaka et al[42]	2018	US	Single center	2005-2014	Mixed	Descending
Acher et al[43]	2019	US	Single center	1984-2014	Mixed	Descending
Uehara et al[48]	2020	Japan	Single center	2007-2017	Mixed	Descending
Orelaru et al[49]	2023	US	Single center	1993-2023	Intact	Descending
Gopaldas et al[28]	2010	US	NIS	2006-2007	Intact	Descending
Goodney et al[31]	2011	US	MEDPAR	1998-2007	Mixed	Descending
Von Allmen et al[36]	2014	UK	HES	2006-2011	Mixed	Descending
Ultee et al[41]	2017	US	NIS	2005-2012	Ruptured	Descending
Chiu et al[44]	2019	US	CMS	1999-2010	Intact	Descending
Geisbüsch et al[45]	2019	Germany	GFSO	2005-2014	Mixed	Descending
Shimizu et al[46]	2019	Japan	JCVSD	2015-2016	Mixed	Descending
Yamaguchi et al[47]	2019	Japan	JROAD-DPC	2012-2015	Ruptured	Descending

CMS, Centers for Medicare & Medicaid Services; GFSO, German Federal Statistical Office; HES, Hospital Episode Statistics; JCVSD, Japan Cardiovascular Surgery Database; JROAD-DPC, The Japanese Registry of all Cardiac and Vascular Diseases Diagnostic Procedure Combination; MEDPAR, Medicare Provider Analysis and Review; NIS, Nationwide Inpatient Sample.

According to the quality assessment based on the NOS, all included studies were rated as high quality (Supplementary Table 3, available at: http://links.lww.com/JS9/D775).

Short-term outcomes

Postoperative mortality

Of the 29 included studies, 28 studies comprising 44 157 patients reported postoperative mortality data. The pooled outcome revealed that the postoperative mortality in the endovascular group was significantly lower than that in the open repair group (OR: 0.57, 95% CI: 0.45-0.72; I² = 72.58%; Fig. 2). In addition, analyses according to different aneurysm types were performed. In the intact DTAA subgroup, 16 studies reported mortality information, with a significantly lower mortality observed in the endovascular group compared with that in the open group (OR: 0.58, 95% CI: 0.38-0.88; I² = 83.34%; Fig. 3A). Similarly, in the rupture DTAA subgroup, a pooled outcome of 7 studies demonstrated that endovascular repair also significantly decreased mortality (OR: 0.54, 95% CI: 0.48-0.61; I² = 0%; Fig. 3B).

Figure 2.

Forest plots for the comparison between endovascular repair group and open repair group for treatment of patients with descending thoracic aortic aneurysm in terms of postoperative mortality.

Figure 3.

Forest plots for the comparison between endovascular repair group and open repair group for treatment of patients with intact descending thoracic aortic aneurysm (A) and rupture descending thoracic aortic aneurysm (B) in terms of postoperative mortality.

Operation details

Fifteen studies including 14 548 patients with DTAA reported data regarding the length of hospital stay. Endovascular repair was associated with a shorter hospital stay (WMD: −6.74 days, 95% CI: −8.61 to −4.87; I² = 90.25%; Supplementary Figure 1, available at: http://links.lww.com/JS9/D774). Nine studies including 1609 patients reported data regarding the length of ICU stay. Endovascular repair was associated with a shorter ICU stay (WMD: −4.13 days, 95% CI: −5.36 to −2.90; I² = 59.44%; Supplementary Figure 2A, available at: http://links.lww.com/JS9/D774). In addition, a significantly shorter length of intervention was also observed in the endovascular group compared with the open group (WMD: −157.79 days, 95% CI: −223.57 to −92.01; I² = 93.54%; Supplementary Figure 2B, available at: http://links.lww.com/JS9/D774).

Short-term postoperative morbidity

Postoperative complications were compared between the endovascular and open surgery groups. Pooled outcomes revealed that patients who underwent endovascular repair exhibited a significant lower risk for cardiac complications (OR: 0.43, 95% CI: 0.26-0.72; I² = 70.40%; Supplementary Figure 3, available at: http://links.lww.com/JS9/D774), paraplegia or spinal cord ischemia (OR: 0.62, 95% CI: 0.50-0.77; I² = 2.52%; Supplementary Figure 4, available at: http://links.lww.com/JS9/D774), pulmonary complications (OR: 0.29, 95% CI: 0.20-0.41; I² = 77.96%; Supplementary Figure 5, available at: http://links.lww.com/JS9/D774), renal complications (OR: 0.35, 95% CI: 0.23-0.55; I² = 90.47%; Supplementary Figure 6, available at: http://links.lww.com/JS9/D774), and stroke (OR: 0.71, 95% CI: 0.54-0.93; I² = 34.69%; Supplementary Figure 7, available at: http://links.lww.com/JS9/D774) than those received open repair.

Sensitivity analysis, meta-regression, publication bias and certainty assessment

Sensitivity analyses were performed to assess the robustness of the model in patients with DTAA and those with intact DTAA. After omitting one study at a time, the results revealed that the pooled mortality outcomes were consistent and without apparent fluctuations in both the DTAA and intact DTAA patient populations (Fig. 4). Egger’s tests comparing endovascular repair with open repair in terms of pooled mortality outcomes in patients with DTAA (P = 0.143) and intact DTAA (P = 0.468) indicated no potential publication bias (Fig. 5). However, when pooling the mortality outcomes of the DTAA and intact DTAA groups, high heterogeneity was observed. To explore the sources of heterogeneity, multivariate meta-regression was performed; however, no significant factors influencing the pooled OR on mortality were detected (Supplementary Table 4, available at: http://links.lww.com/JS9/D775). The GRADE levels of evidence for mortality in the DTAA patient group for postoperative paraplegia or spinal cord ischemia, postoperative pulmonary complications, postoperative stroke, postoperative renal complications, and DTAA rupture were low, whereas the levels of evidence for the other items were all very low (Supplementary Table 5, available at: http://links.lww.com/JS9/D775).

Figure 4.

Sensitivity analysis plot based on postoperative mortality in comparison between endovascular repair group and open repair group for treatment of patients with descending thoracic aortic aneurysm (A) and intact descending thoracic aortic aneurysm (B).

Figure 5.

The Egger’s test for postoperative mortality in terms of postoperative mortality in comparison between endovascular repair group and open repair group for treatment of patients with descending thoracic aortic aneurysm (A) and intact descending thoracic aortic aneurysm (B).

Long-term outcomes

All-cause survival

In the estimated reconstructed survival dataset, because only a few patients had a follow-up period beyond 132 months, especially in the endovascular repair group, the follow-up was limited to 132 months to preserve statistical power. Nineteen studies involving 22 933 patients with DTAA (endovascular, n = 7479; open, n = 15 454) reported data regarding all-cause survival. The pooled estimated KM curves for the all-cause survival is presented in Fig. 6A, and demonstrated that patients who received open repair exhibited significantly longer all-cause survival than those who underwent endovascular repair (HR: 1.20, 95% CI: 1.15-1.26, P < 0.001). However, through visual inspection, the KM curves intersected around the 16-month time point. The Schoenfeld residuals plot and Grambsch-Therneau test (P < 0.001; Supplementary Figure 8A, available at: http://links.lww.com/JS9/D774) further emphasized that the cohort seriously violated the proportional hazards assumption. Analysis of the time-varying HRs for mortality based on flexible parametric survival models with B-splines revealed that endovascular repair was favored in the early stages (Fig. 6B). Subsequently, landmark analysis was performed and 16 months (KM curves crossing the time point) was designated as the landmark time point. In the first 16 months after intervention, the landmark analysis revealed a significantly lower risk for mortality in the endovascular repair group (HR: 0.80, 95% CI: 0.75-0.86, P = 0.021). However, the landmark analysis beyond 16 months yielded a significant reversal of the HR favoring open repair (HR: 1.64, 95% CI: 1.53-1.75, P < 0.001; Fig. 6C). In the RMST analysis, the early survival benefit was attributed to a reduced mortality risk in the endovascular arm, and open repair was associated with longer mean all-cause survival compared with endovascular repair from the four-year follow-up onward (Table 2).

Figure 6.

Pooled survival curves of all-cause survival comparing endovascular repair with open repair for treatment of patients with DTAA (A) and intact DTAA (E). Time-varying HRs with 95% CI for all-cause survival in endovascular repair compared with open repair for treatment of DTAA (B) and intact DTAA (E) at every given time during follow-up; these are derived from flexible parametric survival models with B‐splines. Landmark analysis of all-cause survival comparing endovascular repair with open repair for treatment of patients with DTAA, designating 16 months of follow‐up as the landmark time (C). Landmark analysis of all-cause survival comparing endovascular repair with open repair for treatment of patients with intact DTAA, designating 7 months of follow‐up as the landmark time (F). DTAA, descending thoracic aortic aneurysm.

Table 2.

Restricted mean survival time analysis of all-cause survival in patients with descending thoracic aortic aneurysm or intact descending thoracic aortic aneurysm treated with endovascular versus open repair

Follow-up	RMST (95% CI), months		RMST difference (95% CI), months	P value	RMST ratio (95% CI), months	P value
	Endovascular	Open
DTAA
6 months	5.516 (5.483 to 5.548)	5.280 (5.251 to 5.309)	0.236 (0.192 to 0.279)	<0.001	1.045 (1.036 to 1.053)	<0.001
1 year	10.671 (10.595 to 10.747)	10.307 (10.246 to 10.369)	0.363 (0.265 to 0.462)	<0.001	1.035 (1.026 to 1.045)	<0.001
2 years	20.335 (20.159 to 20.510)	20.004 (19.871 to 20.137)	0.331 (0.111 to 0.551)	0.003	1.017(1.006 to 1.028)	0.003
3 years	29.145 (28.859 to 29.431)	29.275 (29.068 to 29.483)	−0.131 (−0.484 to 0.223)	0.469	0.996 (0.984 to 1.008)	0.469
4 years	37.132 (36.727 to 37.538)	38.057 (37.770 to 38.344)	−0.925 (−1.422 to −0.428)	<0.001	0.976 (0.963 to 0.989)	<0.001
5 years	44.422 (43.889 to 44.956)	46.283 (45.913 to 46.654)	−1.861 (−2.511 to −1.212)	<0.001	0.960 (0.946 to 0.974)	<0.001
10 years	70.190 (68.895 to 71.485)	77.584 (76.503 to 78.664)	−7.394 (−9.080 to −5.707)	<0.001	0.905 (0.884 to 0.926)	<0.001
Intact DTAA
6 months	5.574 (5.541 to 5.607)	5.442 (5.415 to 5.468)	0.133 (0.090 to 0.175)	<0.001	1.024 (1.016 to 1.032)	<0.001
1 year	10.799 (10.720 to 10.877)	10.656 (10.598 to 10.715)	0.142 (0.044 to 0.241)	0.005	1.013 (1.004 to 1.023)	0.004
2 years	20.604 (20.421 to 20.788)	20.735 (20.607 to 20.863)	−0.130 (−0.354 to 0.093)	0.253	0.994 (0.983 to 1.005)	0.254
3 years	29.531 (29.229 to 29.832)	30.394 (30.191 to 30.597)	−0.863 (−1.227 to −0.500)	<0.001	0.972 (0.960 to 0.984)	<0.001
4 years	37.616 (37.186 to 38.046)	39.562 (39.279 to 39.845)	−1.946 (−2.461 to −1.431)	<0.001	0.951 (0.938 to 0.964)	<0.001
5 years	44.959 (44.390 to 45.527)	48.166 (47.797 to 48.535)	−3.207 (−3.885 to −2.529)	<0.001	0.933 (0.920 to 0.947)	<0.001
10 years	70.415 (69.013 to 71.817)	80.096 (78.901 to 81.292)	−9.681 (−11.524 to −7.839)	<0.001	0.879 (0.858 to 0.901)	<0.001

CI, confidence interval; DTAA, descending thoracic aortic aneurysm; RMST, restricted mean survival time.

In addition, patients with intact DTAA were also analyzed. In total, 20 221 patients (endovascular, n = 6351; open, n = 13 870) from 12 studies were evaluated for all-cause survival. The pooled KM curves demonstrated that endovascular repair was associated with a significantly higher risk for long-term mortality (HR: 1.34, 95% CI: 1.27-1.41, P < 0.001; Fig. 6D). Similarly, visual evidence of violation of the proportional hazards assumption in the KM curves (approximately the seven-month time point) was apparent and supported by the Schoenfeld residuals plot and Grambsch-Therneau test (P < 0.001; Supplementary Figure 8B, available at: http://links.lww.com/JS9/D774).The analysis of time-varying HRs for mortality based on flexible parametric survival models with B-splines revealed that endovascular repair was favored in early stage (Fig. 6E). The landmark analysis revealed that endovascular repair was associated with significant lower risk of mortality in the first 7 months after intervention (HR: 0.80, 95% CI: 0.73-0.88, P = 0.024), but yielded a significant reversal of the HR favoring open repair beyond 7 months (HR: 1.72, 95% CI: 1.61-1.84, P < 0.001; Fig. 6F). In the RMST analysis, the early survival benefit was attributed to reduced mortality risk in the endovascular arm, and open repair was associated with longer mean all-cause survival than endovascular repair from the three-year follow-up (Table 2).

Freedom from aortic-related survival

Six studies involving 1793 patients (endovascular, n = 1054; open, n = 739) with DTAA reported data regarding freedom from aortic-related survival. Pooled results revealed that patients in endovascular repair group exhibited significantly longer freedom from aortic-related survival (HR: 0.71, 95% CI: 0.54-0.93, P = 0.012; Fig. 7A).

Figure 7.

Pooled survival curves of freedom from aortic-related survival (A) and freedom from aortic-related reintervention (B) comparing endovascular repair with open repair for treatment of patients with descending thoracic aortic aneurysm.

Freedom from aortic-related reintervention

Similarly, in the estimated reconstructed survival dataset, because no patients in the endovascular repair group reported a follow-up beyond 132 months, the follow-up was limited to 132 months to preserve statistical power. Five studies involving 1368 patients (endovascular, n = 915; open, n = 453) with DTAA reported data regarding freedom from aortic-related reintervention. The pooled results demonstrated that the freedom from aortic-related survival was significant higher in the open repair group (HR: 2.10, 95% CI: 1.45-3.04, P < 0.001; Fig. 7B).

Discussion

The present systematic review and meta-analysis comprehensively compared short- and long-term clinical outcomes between endovascular repair versus open repair in the treatment of patients with DTAA. Results revealed that endovascular repair was associated with a lower risk for postoperative mortality, morbidity, and freedom from aortic-related survival but a higher rate of all-cause survival and freedom from aortic-related reintervention than those for open repair.

Several previous meta-analyses have compared endovascular and open repairs in patients with thoracic aortic diseases^[10-12,51]. Our meta-analysis focused only on patients with DTAA, and, for the first time, we reconstructed individual patient data for long-term survival. Alsawas et al and Harky et al^[10,11] compared the short-term outcomes of the two interventions and found a significant superiority of endovascular repair in terms of perioperative outcomes. However, these two meta-analyses were published in 2017 and 2019, respectively, and the publication year of the included studies was up to 2015. In addition, the two meta-analyses lacked information regarding long-term survival outcomes. Therefore, evidence from these two meta-analyses may lack timeliness and representativeness.^[10,11]. Mahboub-Ahari et al^[51] included 15 studies, and their results were broadly similar to those in our meta-analysis. However, although the research participants in the study by Mahboub-Ahari et al^[51]. were diagnosed with thoracic aortic aneurysm (TAA), which included DTAA and covered a wider range, the study included fewer studies than our meta-analysis. In addition, Mahboub-Ahari et al^[51] also included studies involving patients with ascending or arch TAA or non-aneurysm thoracic aortic diseases (including type B aortic dissection and traumatic aortic disease). Therefore, the literature search strategies and definitions of the eligibility criteria in the study by Mahboub-Ahari et al^[51] may differ significantly from those of the present meta-analysis, reasonably leading to different conclusions. In another meta-analysis, Iyanna et al^[12] included four propensity score matching studies and four studies reporting risk-adjusted outcomes and revealed that open repair was associated with better long-term survival but found no significant difference in terms of operative mortality between the two interventions, which was different from our results. Although their meta-analysis included high-quality studies that adjusted for potential influencing factors, the small number of included studies may have resulted in the omission of many meaningful studies, and individual large sample size studies may have had a high weight, thus affecting the pooled results. On the other hand, Iyanna et al^[12] reported survival outcomes at different time periods in segments, and only 2 studies reported survival data with follow-up beyond two years. The present meta-analysis included sufficient survival data and was reconstructed to enable more intuitive observation and comparison of long-term survival changes between the two different intervention groups. In addition, among the included studies, the research participants were patients diagnosed with TAA, which may have included those with arch and ascending TAA who did not fulfill the inclusion criteria for their meta-analysis^[52].

Since TEVAR was approved in 2005, it has become an alternative option to open repair and has even largely supplanted conventional open repair due to its favorable outcomes, ease of use, and short duration of hospitalization^[4,29,53]. A study including 9518 TEVARs from 13 countries based on the VASCUNET trial reported that TEVAR has become the primary surgical treatment modality for descending aortic pathology^[54]. Another observational study from England, spanning a 23-year period, observed that the increasing trend in surgical interventions for patients with TAA was mainly attributable to the increase in TEVAR^[55]. Before the introduction of TEVAR, some patients with advanced age and comorbidities may have lost the opportunity to undergo surgery due to the lack of availability of open surgical repair. In addition, an observational study revealed that TEVAR was associated with a lower risk for aortic-related mortality than open repair, regardless of age, sex, or aneurysm rupture status^[55]. In fact, although the endovascular repair group in the vast majority of studies included in the present analysis had a mean age older than that of the open repair group, the pooled outcomes still demonstrated the significant superiority of endovascular repair over open repair in terms of postoperative mortality, regardless of aneurysm rupture or intact status. However, high heterogeneity in postoperative mortality was observed. All studies included in our meta-analysis were retrospective in design and had long follow-up periods. Therefore, there may have been significant differences in the devices used, type of stent, postoperative care, spinal cord protection protocols, population characteristics, and other aspects within and among the included studies. After performing regression analysis, sources of heterogeneity remained elusive; as such, large-scale randomized controlled trials are needed to verify these results.

Concerning safety outcomes, we found a lower risk for paraplegia or spinal cord ischemia, and cardiac, pulmonary, and renal complications in the endovascular repair group. Spinal cord injury is mainly caused by occlusion of the intercostal and lumbar arteries by an aortic graft or endovascular stent, perioperative hypotension, or embolic events^[1]. The incidence of permanent spinal cord ischemia is reportedly 2%–10%^[44,56,57]. Previous meta-analyses that included a large number of studies revealed that open repair was associated with a higher incidence of spinal cord injury than endovascular repair, which is consistent with our results^[58,59]. Interestingly, Awad et al^[60] compared the pathophysiology of spinal cord ischemia after endovascular and open repair by through a newly developed endovascular repair dog model and found the mechanisms of spinal cord injury from endovascular repair were drastically different from those from open repair. White matter edema in L3–L5 localized to the dorsal column medial lemniscus area associated with loss of myelin basic protein but not neurons after endovascular repair, versus massive neuronal loss in the gray matter in L3–L5 after open repair^[60]. Fortunately, cerebrospinal fluid drainage may be an effective measure for reducing the occurrence of postoperative spinal cord ischemia^[61,62], and was recommended by American Heart Association /American College of Cardiology^[63]. Postoperative stroke is closely related to hemodynamic changes in patients before and after surgery, and is one of the main causes of postoperative disability and death. During endovascular repair procedures, atheromatous plaque debris caused by guidewires, catheters, or deployment of stent grafts may lead to stroke^[64]. Based on the higher postoperative survival rate of patients, a lower incidence of postoperative stroke indicates better late survival and quality of life^[65-67]. In addition, compared with the advantages of low-level trauma and simple endovascular repair, hypovolemia, extracorporeal circulatory support, intraoperative bleeding, thoracotomy, use of aortic clamps, prolonged mechanical ventilation and ICU stay, hemodynamic disturbance, myocardial injury, and severe myocardial stress in open repair result in a higher risk for renal, pulmonary, and cardiac complications^[68-72]. Patients may take longer to return to their preoperative level of health-related quality of life after open repair^[73].

Endovascular repair has demonstrated significant postoperative advantages over conventional open repair in the treatment of TAA, especially thoracic aortic rupture^[74,75]. However, it is also associated with a higher risk for reintervention due to endoleaks, retrograde type A aortic dissection, endograft migration, and aneurysm rupture^[76-79]. Although the indications of endovascular repair have grown to younger and low-risk patients, the need for appropriate preoperative aortic anatomy and the higher rate of reinterventions may present barriers to wider adoption^[80].

Therefore, the long-term survival efficacy of open and endovascular repairs remains controversial. However, although the development of TEVAR has been rapid since its introduction and has become an alternative option to open repair, with improvements in overall medical technologies and the maturation of scientific perioperative management, open repair may also achieve good clinical outcomes. In our meta-analysis, we observed that TEVAR only provided short- to long-term all-cause survival benefits for patients with TAA or DTAA, whereas open repair provided a significant reversal of long-term all-cause survival benefits. These results are consistent with those of a meta-analysis by Iyanna et al^[12]. Another meta-analysis of reconstructed IPD reported that endovascular repair offered a better survival advantage in the first 11 months, whereas open repair provided a better survival outcome after 11-month time point and up to 180 months for patients with uncomplicated abdominal aortic aneurysms^[81]. Although open repair causes a higher risk for perioperative complications, affecting the short-term survival outcomes of patients, it may provide more definitive surgical outcomes, resulting in better long-term survival. In addition, endovascular repair was associated with worse freedom from aortic-related re-intervention in our study. Although our results revealed a lower risk for postoperative complications in the endovascular repair group, long-term postoperative adverse outcomes, including aortic aneurysmal dilatation, endoleak, and other related complications, may contribute to a higher risk for re-intervention and even worse long-term survival compared with open repair. However, few meta-analyses have reported the pooled results for aortic-related survival. The KM curves in our meta-analysis presented better freedom from aortic-related survival in the endovascular repair group than that in the open repair group; however, the trend of the KM curves tended to stabilize, which may be because aortic-related events were mainly concentrated in the early and mid-term periods after repair. Long-term secondary aneurysm sac rupture after endovascular repair may be a significant reason for the dilution of aortic-related survival advantage^[82]. Our study demonstrated that an increase in the re-intervention rate did not increase the risk for aortic-related mortality. This appears to indicate that the superiority of short-term survival in endovascular repair can be attributed to the low risk for short-term aortic-related mortality. Over time, the trend in reducing all-cause mortality by decreasing aortic-related mortality in the endovascular repair group gradually diluted and reversed. However, as mentioned above, the impact of differences in patient age and preoperative comorbidities between the groups on the results cannot be ignored. A meta-analysis investigating abdominal aortic aneurysm repair in young patients found no significant difference in overall and aneurysm-related mortality between the two groups^[83]. In addition, in multiple randomized controlled trials and adjusted studies comparing endovascular and open repair in patients with abdominal aortic aneurysms, no significant difference in overall survival was found between the two groups^[84-87]. Two included studies also revealed no significant difference in OS between the 2 groups after propensity score matching^[49,50]; however, pooled results were not performed due to the limited sample size and number of studies. Therefore, unadjusted pooled results should be interpreted with caution. Open repair has historically been considered the gold standard for treatment of TAA, which has well described long-term outcomes as the present study showed. However, patients with advanced age and severe medical comorbidities may only chose to receive endovascular repair that the high risk of open repair is unacceptable. As for special patient populations such as those with connective tissue diseases, open repair needed to be considered. However, studies on selection of treatment strategy for young and low-risk patients with DTAA were limited, so more randomized controlled studies are needed for further exploration in the future.

The present study had several limitations, the first of which was the high level of heterogeneity in our analyses due to the inclusion of non-randomized studies and the limitations of different regions, races, time periods, and intervention details of the included studies. Second, most patients in the endovascular repair group were older, with advanced age, and more severe complications, which may have affected the pooled outcomes, potentially underestimating the long-term survival outcomes of the endovascular repair group.

Conclusion

In patients with DTAA and intact DTAA, open repair was associated with significantly better long-term all-cause survival but a higher risk for postoperative mortality. In the treatment of patients with DTAA, there was longer freedom from aortic-related survival, lower morbidity, and shorter lengths of intervention, ICU stay, and hospital stay in the endovascular group. However, open repair yielded significantly better freedom from aortic-related re-intervention than endovascular repair. Nevertheless, further large-scale, multicenter, randomized controlled studies are required to confirm our findings.

Ethical approval

Ethics approval was not required for this systematic review.

Consent

Informed consent was not required for this systematic review.

Sources of funding

All the authors declare to have received no financial support or sponsorship for this study.

Author’s contribution

Conceptualization: J.L., D.G., S.L. and G.Z.; data curation: J.L., D.G., K.X., Z.L., and P.L.; formal analysis and methodology: K.X., Z.L., P.L., Y.L., Y.W. and Y.Y.; supervision: S.L. and G.Z.; writing – original draft: J.L. and D.G.; writing – review & editing: J.L., D.G., S.L. and G.Z.

Conflicts of interest disclosure

The authors have no relevant financial or non-financial interests to disclose.

Research registration unique identifying number (UIN)

The present study has been registered on PROSPERO (ID: CRD42024590468).

Guarantor

Junning Liu, Dan Gou, Kanglin Xu, Ziao Lu, Peidong Li, Yong Lei, Yongjie Wang, Yuting Yang, Shiqiang Liu and Guiying Zhu accept full responsibility for the work. Junning Liu and Guiying Zhu had access to the data.

Provenance and peer review

This paper was not invited.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Assistance with the study

None.

Presentation

None.