Thoracic endovascular vs open surgical repair in descending thoracic aortic aneurysms: A systematic review and meta-analysis

Abstract

BACKGROUND

Descending thoracic aortic aneurysms are dangerous and have to be treated quickly. The primary treatment methods are thoracic endovascular aortic repair (TEVAR) and open surgical repair (OSR). The comparative effectiveness and safety of TEVAR and OSR were evaluated in this meta-analysis, focusing on perioperative and long-term outcomes.

AIM

To compare and contrast the efficacy and safety of TEVAR vs OSR in the treatment of descending thoracic aortic aneurysms. This study aims to assess both perioperative and long-term outcomes through a systematic review and meta-analysis.

METHODS

A comprehensive search of PubMed, EMBASE, and Cochrane was conducted from inception to January 2025. Baseline characteristics and outcomes were evaluated. Odds ratios (OR) for dichotomous data and mean differences for continuous data with 95% confidence intervals (CI) were analyzed using random-effects models.

RESULTS

A meta-analysis of 21 studies involving 29465 patients (8261 TEVAR; 21204 OR) showed TEVAR associated with lower operative mortality (OR = 0.60, 95%CI: 0.42-0.85, P = 0.004), shorter intensive care unit (-2.94 days, 95%CI: -4.76 to -1.12, P = 0.002) and hospital stays (-7.35 days, 95%CI: -10.54 to -4.17, P < 0.00001), and reduced rates of paraplegia (OR = 0.44, 95%CI: 0.27-0.73, P = 0.002), spinal ischemia (OR = 0.30, 95%CI: 0.16-0.56, P = 0.0002), renal failure (OR = 0.29, 95%CI: 0.14-0.61, P = 0.001), and wound infections (OR = 0.28, 95%CI: 0.13-0.61, P = 0.001). However, TEVAR had higher rates of vascular complications. No significant differences were noted in 1-year and 5-year mortality rates, the rate of non-elective surgery, neurological complications, or stroke rates.

CONCLUSION

Compared to EVAR, TEVAR revealed lower operative mortality and better perioperative outcomes across all indicators, including hospital and intensive care unit stays, as well as fewer complications, except for those related to vascular problems. Mortality results were also similar in the long run; consequently, more research is required concerning the long-term durability.

Key Words: Thoracic endovascular aortic repair; Open surgical repair; Descending thoracic aortic aneurysm; Meta-analysis

Core Tip: Thoracic endovascular aortic repair may be the preferred strategy for many patients, particularly those with significant comorbidities or high surgical risk; treatment decisions should be carefully individualized. A comprehensive assessment of patient-specific factors, including anatomical suitability, overall health status, and institutional expertise, remains critical in determining whether an endovascular or open surgical approach is most appropriate. As technological advances and surgical techniques continue to refine, further studies will be essential in guiding optimal patient selection and improving long-term outcomes for descending thoracic aortic aneurysm repair.

INTRODUCTION

The global incidence of thoracic aortic aneurysms (TAA) is around 6 per 100000 individuals annually[1]; in the United States, the rate is even higher at 10.4 per 100000[2]. Among these, descending TAA (DTAA) account for nearly one-third of all TAAs. Diagnosing DTAA is challenging due to its complex anatomical location; however, advancements in imaging techniques, such as cardiac magnetic resonance imaging, cardiac computed tomography, and cardiac ultrasound, have facilitated the early detection of asymptomatic aneurysms. Early diagnosis is crucial as it prevents progression to aortic dissection and rupture, which have mortality rates approaching 100%.

Management strategies for DTAA involve regular surveillance using cardiac computed tomography scans or magnetic resonance imaging, medical management, and surgical vs endovascular intervention when necessary. Surgical repair options include open surgery, endovascular approaches (utilizing branched or fenestrated grafts), or a hybrid method[3]. The choice between open and endovascular methods is complex, influenced by the aneurysm’s etiology, precise location, and the patient’s age and comorbidities, all of which can affect both immediate and long-term survival.

Open surgical repair (OSR) was the default standard of care for the treatment of DTAA, primarily due to its positive impact on life expectancy (over 10 years)[4], and its demonstrated durability[5]. However, studies have revealed its association with an increased risk of immediate and short-term morbidities, including paraplegia and spinal cord paraparesis/spinal cord injury, renal failure with dialysis dependence, stroke, and life-threatening gastrointestinal bleeding (infrequent)[6].

Thus, especially in the last decade, the thoracic endovascular aortic repair (TEVAR) approach has been increasingly offered to a greater pool of eligible patients. Systematic reviews of observational studies have shown better 30-day mortality and morbidity outcomes with TEVAR compared to OSR[7-9].

However, long-term durability remains a significant deterrent, with secondary open operation rates[10] of subsequent endovascular interventions following thoracic stent-graft procedures as high as 32 percent[10]. Harky et al[11] reported that TEAVR resulted in better perioperative outcomes and reduced length of stay. However, there was no significant difference in long-term outcomes, including one-year and five-year mortality rates. These findings corroborate the results of an earlier meta-analysis by Cheng et al[8] in 2010. Moreover, over time, there have been significant advances in the OSR approach, from developments in organ preservation to advancements in neuroprotective methods, resulting in improved outcomes for morbidity and mortality[12].

Important limitations of the current literature include the results reported from observational comparative studies vs randomized clinical trials. In conclusion, our meta-analysis aims to fill important clinical gaps in the existing evidence base regarding the comparative outcomes of TEVAR vs OSR approaches, particularly concerning long-term outcomes.

MATERIALS AND METHODS

Study design

This systematic review and meta-analysis were conducted to compare the outcomes of TEVAR and OSR for DTAA. The review adheres to the guidelines outlined in the PRISMA statement for reporting systematic reviews and meta-analyses[13].

Eligibility criteria

We included observational studies that directly compared TEVAR and OSR in the treatment of DTAA. Studies must have reported on key outcomes, including operative mortality, complications, and long-term survival (1-year and 5-year mortality rates). Studies focusing on ruptured aneurysms, pediatric populations, or those lacking comparative data between TEVAR and OSR were excluded.

Information sources and search strategy

We systematically searched four major databases: PubMed, EMBASE, Cochrane, and Ovid. The search was conducted using terms such as “descending thoracic aortic aneurysm”, “open surgical repair”, “endovascular repair”, and “thoracic aortic stenting” in various combinations, including Medical Subject Heading terms where applicable. The search spanned from database inception to January 2025. No language or publication date restrictions were applied. We also performed manual reference checks to identify additional relevant studies.

Study selection

Two independent reviewers (Ali U and Khawar M) screened the studies for eligibility. Discrepancies between reviewers were resolved by consensus or by involving a third reviewer. Studies that met the inclusion criteria were then assessed for quality and relevance, and data were extracted accordingly. A flowchart illustrating the study selection process was constructed by the PRISMA guidelines.

Data extraction

Data was independently extracted by three reviewers (Ali U, Pathak P, Mubarika M) using a pre-defined extraction form. Extracted data included: Study characteristics (e.g., authors, year, country, study design), Patient demographics (e.g., age, gender, comorbidities), Surgical outcomes (e.g., operative mortality, complications such as stroke, paraplegia, renal failure, wound infection), Follow-up duration and long-term mortality data (1-year and 5-year outcomes). Any discrepancies in data extraction were resolved through discussion, with the assistance of a fourth reviewer (Saifullah M) when necessary.

Quality assessment

Two independent reviewers used the Newcastle-Ottawa Scale to assess the methodological quality of the included studies. The scale evaluates three domains: Selection (4 items), comparability (1 item, up to 2 stars), and outcome assessment (3 items), with a maximum possible score of 9 stars. Discrepancies were resolved through discussion with a third reviewer[14].

Statistical analysis

Data were synthesized using random-effects meta-analysis to account for the expected heterogeneity across studies. The following statistical measures were calculated: Odds ratios (OR) for dichotomous outcomes (e.g., mortality, complications). Mean differences for continuous outcomes [e.g., length of intensive care unit (ICU) stay, hospital stay]. Both ORs and mean differences were accompanied by 95% confidence intervals (CIs). Heterogeneity among studies was assessed using the I² statistic, with values greater than 50% indicating substantial heterogeneity. Sensitivity analyses, including leave-one-out methods, were conducted to assess the impact of individual studies on heterogeneity and model selection.

To assess potential publication bias, we performed Egger’s regression test on studies that included at least 10 participants. This test evaluated the asymmetry of the funnel plot for each key outcome and was used to determine if there was any evidence of publication bias. All statistical analyses were conducted using RevMan 5.3 (Cochrane Collaboration). The Egger’s regression test for publication bias was performed using R version 4.4.1. Following the statistical analysis, the quality of evidence for each outcome was evaluated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach, which assesses domains such as risk of bias, inconsistency, imprecision, indirectness, and effect size to determine the certainty of the evidence.

RESULTS

Included studies

The PRISMA statement flowchart (Figure 1) outlines the literature screening process, study selection, and exclusion criteria. The initial search yielded 1937 articles, from which 46 full-text articles were retrieved for assessment. Ultimately, 21 studies[15-35] met the eligibility criteria and were included in both the qualitative and quantitative meta-analyses.

Figure 1  PRISMA flowchart outlining the literature screening process, study selection, and exclusion criteria.

Study characteristics

Studies spanned from 2004 to 2023 and were conducted in the United States, Germany, Switzerland, Denmark, the Netherlands, France, and Korea. Sample sizes ranged from 22 patients to 11565 patients, with follow-up durations ranging from 30 days to 9 years. The mean age for TEVAR patients ranged from 64 ± 14.07 years to 80.6 ± 4.0 years, while for OSR patients, it ranged from 56.4 ± 14.3 years to 76.8 ± 1.8 years. Male participants made up 40% to 84% of the samples. TEVAR patients were generally older and had more comorbidities (e.g., hypertension, coronary artery disease, diabetes, renal dysfunction), indicating a sicker population. Detailed characteristics of each study are presented in Table 1.

Table 1 Detailed characteristics of each included study (e.g., sample size, follow-up duration, mean age, comorbidities).

Ref.	Study design	Country	Primary endpoints	No. of patients		Follow up duration		Age		Male (%)		Hypertension		CAD		DM
				EVAR	Open	EVAR	Open	EVAR	Open	EVAR	Open	EVAR	Open	EVAR	Open	EVAR	Open
von Allmen et al[16], 2014	Observational study	United Kingdom	30 days operative mortality, long-term survival (5 years), aortic related intervention	354	264	5 years (maximum)	5 years	72.2 ± 3.5	70.3 ± 3.8	65.5	51.9	NA	NA	NA	NA	NA	NA
Andrassy et al[17], 2011	Observational study	Germany	Peri-operative morbidity and mortality, 1 year mortality and cumulative long term survival	53	24	34 ± 35 (mean)	53 ± 55	70 ± 10	67 ± 12	60	54	85%	83%	40%	33%	12%	7%
Arnaoutakis et al[18], 2015	Observational study	United States	1 year all-cause mortality	62	56	23.7 months (median)	36.4 months	67.6 ± 12.9	56.4 ± 14.3	53	68	89	77	34	9	16	11
Bavaria et al[19], 2007	Observational study	United States	Peri operative mortality and complications, middle term survival and reoperation rates	140	94	25.8 ± 14.6 (mean)	24.9 ± 12.8	70.5 ± 10.4	68.2 ± 10.2	57	51	-	-	49	36	-	-
Brandt et al[20], 2004	Observational study	Germany	30 day mortality, peri operative morbidity	22	22	30 days	30 days	68 ± 13	69 ± 11	68	59	64	77	36	32	14	18
Chiu et al[28], 2019	Observational study	United States	All-cause mortality	2470	1235	4.53 ± 2.89	5.43 ± 6.90	72.97 ± 8.36	72.88 ± 7.52	57.6	57.5	78.9	79.5	-	-	21.5	19.4
Desai et al[44], 2012	Observational study	United States	Operative mortality, late survival	106	45	59 ± 35.32	81 ± 71.99	74.3 ± 9.1	69.5 ± 11.1	57	40	88	89			21	6.7
Dick et al[21], 2008	Observational study	Switzerland	Peri operative mortality and morbidity within 30 days of treatment, cumulative long term survival, quality of life	52	70	29 ± 16	37 ± 17	68.8 ± 10.1	61.6 ± 15.2	83	84	79	76	35	49	15	4
Glade et al[22], 2005	Observational study	Netherlands	In-hospital mortality, middle term survival	42	53	15	26	67	67	62.1	62.1	86	75	-	-	12	6
Goodney et al[45], 2011	Observational study	United States	-	2433	11565	-	-	75.9 ± 6.29	73.8 ± 5.49	58.7 (95% CI: 567-60.7)	55.4	-	-	-	-	NA	NA
Gopaldas et al[23], 2010	Observational study	United States	In-hospital mortality	2563	9106	NA	NA	69.5 ± 12.7	60.2 ± 14.2	59.4	68	-	-	-	-	13.7	9.4
Hughes et al[30], 2014	Observational study	United States	Mortality	712	8255	Short term	-	71 ± 11.14	62.33 ± 14.83	61.3	65.6	-	-	-	-	13.2	8.3
Karimi et al[24], 2012	Observational study	United States	In-hospital and late mortality, endo-leak and re-intervention rates	28	29	26.9 ± 14	42.6 ± 19.8	73.2 ± 9.7	62.3 ± 12.2	57.1	65.5	78.6	89.7	32.1	27.6	10.7	34.5
Kieffer et al[31], 2009	Observational study	France	-	52	121	Short term	-	69.3 ± 12.5	59.4 ± 13.7	78.8	82.6	71.2	77.7	34.6	24.8	5.8	4.1
Lee et al[25], 2015	Observational study	Korea	30-day and late mortality	114	53	36 ± 26	36 ± 26	65.5 ± 12.9	60.1 ± 15.9	76.3	69.8	73.7	75.5	11.4	13.2	8.8	13.2
Matsumura et al[32], 2014	Observational study	Denmark	-	158	70	5 years (maximum)	5 years	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
Ogawa et al[35], 2021	Observational study	United States	30-day and long term all cause mortalities	79	39	828 ± 1258 (days)	1048 ± 1591 (days)	70.4 ± 12.7	64 ± 13.6	60.8	74.4	86.1	69.2	15.2	12.8	8.9	7.7
Orandi et al[26], 2009	Observational study	United States	In-hospital mortality	267	763	NA	NA	69.9 ± 20.9	66.1 ± 21.3	69.2	66.1	76.2	64.3	NA	NA	7.6	11
Orelaru et al[34], 2023	Observational study	United States	-	120	120	8.80 ± 7.13 years	4.0 7 ± 3.68	64 ± 14.07	65 ± 10.37	64	66	91	88	18	25	13	13
Patel et al[33], 2008	Observational study	United States	-	52	41	33.1 ± 36.9 (maximum follow up 12 years)	33.1 ± 36.9	80.6 ± 4.0	76.8 ± 1.8	50	70.7	78.9	80.5	55.8	43.9	13.5	4.9
Stone et al[27], 2006	Observational study	United States	Peri-operative mortality, actuarial middle-term survival, freedom from re-intervention	105	93	22 ± 16.83 months	-	70 ± 14.3	70.8 ± 9.8	62.9	54.8	NA	84.9	NA	NA	NA	NA

EVAR: Endovascular aortic repair; NA: Not available; CI: Confidence interval.

Quality assessment

Of the 21 studies assessed, 19 achieved scores of 8 or 9, indicating high methodological quality with a low risk of bias. Two studies received a score of 7, reflecting moderate quality and a moderate risk of bias. No studies were classified as having a high risk of bias (Table 2).

Table 2 Quality assessment scores of the included studies, indicating methodological quality and risk of bias.

Ref.	Selection				Comparability	Outcome			Total
	1	2	3	4	1	1	2	3
von Allmen et al[16], 2014	One star	One star	One star	One star	One star	One star	One star	One star	8
Andrassy et al[17], 2011	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Arnaoutakis et al[18], 2015	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Bavaria et al[19], 2007	One star	One star	One star	One star	One star	One star	One star	-	7
Brandt et al[20], 2004	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Chiu et al[28], 2019	One star	One star	One star	One star	One star	One star	One star	One star	8
Desai et al[44], 2012	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Dick et al[21], 2008	One star	One star	One star	One star	Two stars	One star	One star	-	8
Glade et al[22], 2005	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Goodney et al[45], 2011	One star	One star	One star	One star	One star	One star	One star	One star	8
Gopaldas et al[23], 2010	One star	One star	One star	One star	One star	One star	One star	One star	8
Hughes et al[30], 2014	One star	One star	One star	One star	-	One star	One star	One star	7
Karimi et al[24], 2012	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Kieffer et al[31], 2009	One star	One star	One star	One star	One star	One star	One star	One star	9
Lee et al[25], 2015	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Matsumura et al[32], 2014	One star	One star	One star	One star	Two stars	One star	One star	-	8
Ogawa et al[35], 2021	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Orandi et al[26], 2009	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Orelaru et al[34],2023	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Patel et al[33], 2008	One star	One star	One star	One star	Two stars	One star	One star	One star	9
Stone et al[27], 2006	One star	One star	One star	One star	Two stars	One star	One star	One star	9

The stars included in the table represent the scores assigned using the Newcastle-Ottawa Scale, a tool designed to assess the quality of non-randomized studies in systematic reviews. Each star corresponds to a point awarded based on three key domains: The selection of study groups, the comparability of the groups, and the ascertainment of exposure or outcome. The maximum score is nine stars, with higher numbers indicating higher study quality. The stars in the table reflect the evaluation of each study against these criteria.

Clinical outcomes comparison

Our study evaluated the following clinical outcomes for TEVAR and OSR: Neurological complications, renal failure, vascular complications, spinal ischemia and wound infection, length of ICU and hospital stay, non-elective surgery, operative mortality, and long-term mortality.

In-hospital results

The in-hospital outcomes were examined, including neurological and renal complications, as well as other perioperative parameters.

Neurological complications: The rate of paraplegia was significantly higher in the OSR group compared to the TEVAR group (OR = 0.44, 95%CI: 0.27-0.73, P = 0.002, I² = 0%) (Figure 2A). In contrast, the incidence of stroke was similar between the two groups, with no statistically significant difference in event rates (OR = 0.90, 95%CI: 0.63-1.29, P = 0.57, I² = 0%) (Supplementary Figure 1A). Overall, the rate of neurological complications did not differ significantly between the two groups (OR = 0.85, 95%CI: 0.70-1.04, P = 0.12, I² = 0%) (Supplementary Figure 1B).

Figure 2 Forest plot. A: Forest plot comparing the rate of paraplegia between thoracic endovascular aortic repair (TEVAR) and open surgical repair (OSR) groups; B: Forest plot comparing the rate of renal failure between TEVAR and OSR groups; C: Forest plot comparing the rate of vascular complications between TEVAR and OSR groups; D: Forest plot comparing the incidence of spinal ischemia between TEVAR and OSR groups. CI: Confidence interval.

Renal failure: The rate of renal failure, defined as the need for dialysis or a creatinine level greater than 200 μmol/L, was higher in the OSR group (OR = 0.29, 95%CI: 0.14-0.61, P = 0.001, (I² = 73%). The heterogeneity dropped significantly after removing Ogawa et al[35] in 2011 (I² = 23%) (Figure 2B).

Vascular complications: The rate of vascular complications, including access port issues, was significantly higher in the TEVAR group (OR = 2.26, 95%CI: 1.40-3.64, P = 0.0008, I² = 40%) (Figure 2C).

Spinal ischemia and wound infection: The incidence of spinal ischemia (OR = 0.30, 95%CI: 0.16-0.56, P = 0.0002, I² = 35%) (Figure 2D) and wound infection (OR = 0.28, 95%CI: 0.13-0.61, P = 0.001, I² = 0%) were also higher in the OSR group (Figure 3A).

Figure 3 Forest plot. A: Forest plot comparing wound infection between thoracic endovascular aortic repair (TEVAR) and open surgical repair (OSR) groups; B: Forest plot comparing the length of intensive care unit between TEVAR and OSR groups; C: Forest plot comparing the length of hospital stay between TEVAR and OSR groups; D: Forest plot comparing operative mortality between TEVAR and OSR groups. CI: Confidence interval.

Length of ICU and hospital stay: Interestingly, the TEVAR group had significantly shorter ICU stays (-2.94 days, 95%CI: -4.76 to -1.12, P = 0.002, I² = 84%) dropped significantly after removing Matsumura et al[32], in 2014 (I² = 44%) (Figure 3B) and shorter total hospital stays (-7.35 days, 95%CI: -10.54 to -4.17], P < 0.00001, I² = 94%) dropped significantly after removing Ogawa et al[35] (I² = 47%) compared to the OSR group (Figure 3C).

Non-elective surgery: At the one-year mark, the rate of non-elective surgery was higher in the TEVAR group, although the difference was not statistically significant (OR = 1.07, 95%CI: 0.87-1.31, P = 0.51, I² = 50%). The heterogeneity dropped significantly after removing Matsumura et al[32], (I² = 23%) (Supplementary Figure 1C).

Mortality rates

Operative mortality: The operative mortality (defined as death during hospitalization or within 30 days after surgery) was higher in the OSR (OR = 0.60, 95%CI: 0.42-0.85, P = 0.004, I² = 55%), which can likely be attributed to the older age and higher comorbid burden in the TEVAR group. The heterogeneity dropped significantly after removing Matsumura et al[32] (I² = 29%) (Figure 3D).

Long-term mortality: No significant differences in long-term mortality were observed between the two groups at the 1-year (P = 0.42) and 5-year (P = 0.32) follow-up points with considerable heterogeneity of I² = 89% and I² = 76%, suggesting that neither procedure offered a significant survival advantage over the other at these time points. The heterogeneity dropped significantly after removing Matsumura et al[32] (from I² = 89% to 45%) and Kieffer et al[31] (from I² = 76% to I² = 41%) (Supplementary Figure 1D and E).

Publication bias

The results suggest that there is no significant evidence of publication bias for the rate of non-elective surgery, postoperative paraplegia, postoperative stroke, 1-year and 5-year mortality, as the P values exceed the typical significance threshold of 0.05. However, for all neurological complications (P = 0.008), postoperative renal failure (P = 0.0007), postoperative vascular complications (P = 0.0007), and operative mortality (P = 0.0010), the P values are below 0.05, indicating potential publication bias for these outcomes (Figure 3).

GRADE assessment

The GRADE assessment of the meta-analysis, evaluating outcomes of TEVAR vs OSR, revealed varying levels of evidence certainty across 13 outcomes. Operative mortality, renal failure, and vascular complications demonstrated very low certainty due to high heterogeneity (I² = 80%-85%), wide confidence intervals, and significant publication bias (P < 0.05). Neurological complications, stroke, non-elective surgery, and 1-year and 5-year mortality exhibited low certainty, primarily due to imprecision from wide confidence intervals and, for neurological complications, moderate heterogeneity (I² = 60%) and publication bias. In contrast, paraplegia, spinal ischemia, wound infection, and lengths of ICU and hospital stay achieved moderate certainty, supported by low heterogeneity (I² = 0%), narrow confidence intervals, and no publication bias. These findings suggest that while TEVAR is associated with higher risks of certain complications (e.g., paraplegia, renal failure), it offers shorter ICU and hospital stays. However, the evidence is limited by the use of observational study designs and publication bias for several outcomes.

DISCUSSION

This systematic review and meta-analysis reveal important distinctions in perioperative and long-term outcomes between TEVAR and OSR for DTAA. TEVAR demonstrated advantages in the early postoperative period, including lower operative mortality, shorter ICU and hospital stays, and reduced rates of complications such as paraplegia, spinal ischemia, renal failure, and wound infections. These benefits are primarily attributable to its minimally invasive nature, which minimizes physiological stress and accelerates recovery. However, TEVAR was associated with a higher incidence of vascular complications and spinal ischemia, highlighting the procedural complexities and risks of endovascular techniques. The observed perioperative advantages of TEVAR may be influenced by patient selection bias, as TEVAR cohorts were older and had a higher burden of comorbidities (Table 1), which could confound these outcomes. This raises concerns about the generalizability of TEVAR’s benefits, particularly to younger or healthier patients who may be better suited for OSR. Propensity-matched studies were excluded from this meta-analysis due to their limited availability; however, such studies could help mitigate selection bias and provide a more precise comparison of outcomes. While TEVAR offered favorable short-term outcomes, long-term survival at both 1 and 5 years remained comparable to OSR, raising concerns about its durability, particularly given the relatively high rates of secondary interventions (up to 32%). These reinterventions, often necessitated by complications such as endoleaks, graft migration, or aneurysmal degeneration, underscore the need for ongoing surveillance and highlight a critical limitation of TEVAR’s long-term structural integrity. These findings suggest the necessity of individualized treatment planning, considering patient-specific factors such as comorbidities, anatomical suitability, and institutional expertise. Ongoing advancements in endovascular technology, such as branched and fenestrated grafts, may address some durability concerns but could also contribute to heterogeneity in outcomes due to variations in device design, operator experience, and patient anatomy. It is essential to continue research focused on optimizing long-term outcomes[36,37].

The observed reduction in both ICU and total hospital length of stay in patients undergoing TEVAR reflects the benefits of its minimally invasive approach, which results in less physiological stress, decreased perioperative morbidity, and a faster postoperative recovery process[38]. However, despite these notable advantages, TEVAR was associated with a higher incidence of vascular complications, particularly those related to access site issues, as well as an increased risk of spinal ischemia. These findings highlight the technical challenges inherent in endovascular procedures, emphasizing the importance of meticulous preoperative planning, precise procedural execution, and adequate operator expertise to optimize outcomes and minimize complications[39,40]. Given these differences, the choice between TEVAR and OSR should be carefully individualized, taking into account patient-specific factors such as anatomical considerations, comorbidities, and surgical risk. Ongoing advancements in endovascular technology and refinement of technique may help mitigate some of the current limitations of TEVAR, potentially broadening its applicability while improving both short-term and long-term outcomes.

The incidence of paraplegia was found to be significantly lower in patients who underwent OSR compared to those treated with TEVAR. This contrast may be explained by differing pathophysiological mechanisms. Paraplegia in OSR is often linked to prolonged aortic cross-clamping, which causes hypoperfusion injury to the spinal cord due to interrupted blood flow. In contrast, spinal ischemia in TEVAR may result from embolic events, such as micro-emboli dislodged during stent-graft deployment, or from coverage of critical intercostal arteries, leading to segmental spinal cord ischemia. These mechanistic differences underscore the need for tailored neuroprotective strategies, such as cerebrospinal fluid (CSF) drainage or preoperative mapping of the spinal cord vasculature, to mitigate the specific risks associated with each approach. However, when evaluating other neurological outcomes, including stroke and overall neurological complications, no statistically significant differences were observed between the two treatment groups. Although TEVAR is generally associated with a lower risk of spinal cord injury due to its minimally invasive nature, it does not provide a protective effect against other neurological complications, such as stroke.

A recent meta-analysis corroborated these findings, demonstrating that while TEVAR was associated with a reduced incidence of paraplegia, stroke rates remained comparable to those observed in patients undergoing OSR[41]. This highlights the multifaceted nature of neurological outcomes in thoracic aortic aneurysm repair, where different procedural risks must be carefully balanced. The persistence of spinal ischemia as a complication of TEVAR highlights the critical need for ongoing research into neuroprotective strategies aimed at reducing this risk. Strategies such as optimizing spinal cord perfusion, utilizing CSF drainage, and refining procedural techniques may help mitigate neurological complications, particularly in high-risk patients undergoing TEVAR. Given the complexity of these outcomes, further studies are warranted to refine patient selection criteria and develop interventions that enhance the safety and efficacy of TEVAR while minimizing the risk of neurological injury[42].

The long-term outcomes of TEVAR and OSR for DTAA repair were found to be comparable, with no statistically significant differences observed in mortality rates at both the 1-year and 5-year follow-up periods. These findings align with a growing body of evidence from prior studies, which have consistently demonstrated similar overall survival rates between the two approaches over extended follow-up durations[43]. Despite this equivalence in long-term mortality, the durability of TEVAR remains a concern, as secondary surgical interventions were required in up to 32% of patients, often due to complications such as endoleaks, graft migration, or progressive aneurysmal degeneration. This high reintervention rate, particularly in patients with complex anatomical variations (e.g., excessive tortuosity, large aneurysmal diameters greater than 60 mm, or significant aortic calcification), underscores the need for improved device designs and long-term follow-up protocols to enhance the effectiveness of TEVAR[44].

Given the lack of definitive long-term survival superiority of either TEVAR or OSR, these findings emphasize the importance of a patient-centered approach when determining the most appropriate treatment strategy. Factors such as patient age, baseline comorbidities, aneurysm morphology, and overall surgical risk should be carefully considered when selecting the optimal intervention. Advancements in endovascular technology, such as branched and fenestrated grafts, hold promise for improving long-term durability; however, their impact on outcomes may vary due to differences in device availability, operator learning curves, and patient-specific anatomical challenges, which may contribute to the observed heterogeneity in this meta-analysis[45].

Vascular complications were significantly more frequent in patients undergoing TEVAR, with a particularly high incidence of access-related issues. This trend is well-documented in prior research, which has consistently reported a greater occurrence of vascular complications following endovascular procedures compared to OSR[46]. The nature of these complications can vary widely, ranging from minor access site injuries to more severe arterial dissections, perforations, or occlusions that may require additional interventions. Although many of these issues can be effectively managed with endovascular or surgical techniques, they nonetheless represent a significant drawback of TEVAR, especially in patients with complex vascular anatomy or heavily calcified and tortuous access vessels.

The increased risk of vascular complications in TEVAR highlights the critical need for specialized operator training, meticulous preprocedural planning, and the refinement of endovascular techniques to minimize these risks. Emerging strategies, such as the use of smaller-profile delivery systems and percutaneous closure devices, may help reduce access-related complications in the future. The development and implementation of hybrid approaches integrating both open surgical and endovascular methods offer a promising strategy to mitigate these risks, particularly in anatomically challenging cases. As technology continues to evolve, further research is warranted to determine the optimal techniques and patient selection criteria for improving the safety and durability of TEVAR while minimizing associated complications[47,48].

Study limitations and implications for future research

This study has several limitations. One of the most significant concerns is that the included studies were primarily observational, which inherently introduces a risk of bias due to potential confounding variables such as patient age, baseline comorbidities, and aneurysm size. These factors may influence both treatment selection and outcomes, making direct comparisons between TEVAR and OSR less definitive. Random-effects models were employed due to substantial heterogeneity across studies (I² = 80%-85%), particularly for outcomes such as operative mortality and renal failure. Unlike fixed-effects models, they account for both within- and between-study variability, making them suitable when clinical and methodological differences exist. While this approach provides more conservative estimates, it may reduce statistical power and precision. Future analyses could include fixed-effects models in sensitivity analyses for outcomes with lower heterogeneity.

Additionally, the high heterogeneity (I² = 80%-85%) observed for outcomes such as operative mortality, renal failure, and vascular complications represents a notable limitation. This heterogeneity likely reflects variability in patient characteristics (e.g., comorbidities, aneurysm etiology), procedural factors (e.g., operator experience, institutional expertise), and study designs across the included studies. Subgroup or sensitivity analyses to explore these potential sources of heterogeneity were not performed in this meta-analysis due to limited data availability and variability in reporting across studies.

Furthermore, significant publication bias (P < 0.05) was detected for neurological complications, vascular complications, and operative mortality, as shown in Figure 3. This bias may lead to overestimation or underestimation of the actual effect sizes for these outcomes, potentially skewing the conclusions of this meta-analysis. The presence of publication bias suggests that studies with significant or favorable results may have been more likely to be published. In contrast, studies with null or adverse outcomes may be underrepresented. Due to data constraints, adjustments such as the trim-and-fill method were not applied in this analysis. Future meta-analyses should employ such methods to adjust for publication bias and provide a more accurate estimate of effect sizes, enhancing the reliability of conclusions.

Future research should prioritize such analyses to understand the drivers of heterogeneity better and identify patient or procedural factors that may influence outcomes. To address these limitations, well-designed randomized controlled trials are necessary to provide a more accurate comparison between the two approaches and minimize the impact of confounding biases.

Another limitation pertains to the GRADE assessment, where outcomes such as paraplegia and spinal ischemia were assigned a “moderate certainty” rating, partly due to low heterogeneity (I² = 0%). However, this rating may not fully account for known variations in neuroprotective strategies, such as the use of CSF drainage, which can significantly influence the incidence of these complications. The inconsistent application or reporting of such strategies across studies may mask underlying variability in clinical practice, potentially leading to an overestimation of the certainty of these findings. Clinically, this discrepancy is relevant because the effectiveness of TEVAR in reducing spinal cord injury is highly dependent on the use of adjunctive neuroprotective measures, which vary by institution and patient risk profile. Future studies should standardize the reporting of neuroprotective strategies and incorporate their impact into GRADE assessments to better reflect the true certainty of evidence for these outcomes. While this analysis provides valuable insights into perioperative and short-term outcomes, long-term durability remains an area of uncertainty. Specifically, the need for secondary interventions following TEVAR and the long-term structural integrity of endovascular grafts require further investigation. Future research should aim to assess these long-term outcomes more comprehensively, ideally incorporating extended follow-up periods and standardized data collection methods to enhance the reliability of the findings. As surgical and endovascular techniques continue to evolve, it is essential to evaluate the impact of advancements such as branched and fenestrated grafts in TEVAR. As technological advances, such as branched and fenestrated grafts, continue to evolve, their role in reducing heterogeneity and improving outcomes should be rigorously evaluated, particularly given their potential to address anatomical challenges that contribute to complications and reintervention[46-50].

CONCLUSION

Based on the results of this meta-analysis, TEVAR appears to provide superior perioperative outcomes compared to OSR in patients undergoing repair of DTAA. Patients treated with TEVAR experienced shorter ICU and overall hospital stays, along with lower rates of complications such as paraplegia and renal failure. However, this advantage is counterbalanced by a higher incidence of vascular complications and spinal ischemia, highlighting the procedural challenges associated with endovascular repair. Long-term survival outcomes at both 1-year and 5-year follow-up were similar between the two approaches, reinforcing the need for further research into the durability of TEVAR and the factors influencing late reintervention rates.

While TEVAR may be the preferred strategy for many patients, particularly those with significant comorbidities or high surgical risk, treatment decisions should be carefully individualized. A comprehensive assessment of patient-specific factors, including anatomical suitability, overall health status, and institutional expertise, remains critical in determining whether an endovascular or open surgical approach is most appropriate. As technological advances and surgical techniques continue to refine, further studies will be essential in guiding optimal patient selection and improving long-term outcomes for DTAA repair.