Long-Term Outcome of Intact Abdominal Aortic Aneurysm After Endovascular or Open Repair

Abstract

Objective

Endovascular aortic aneurysm repair (EVAR) has been established as a standard treatment option for intact abdominal aortic aneurysm (iAAA) and gained importance due to a lower perioperative mortality than open repair (OAR). However, whether this survival advantage can be maintained or if OAR is beneficial in terms of long-term complications and reinterventions remains questionable.

Design

In this retrospective cohort study data from patients undergoing elective EVAR or OAR for iAAAs in the years 2010-2016 was analyzed. The patients were followed through 2018.

Methods

In the propensity score matched cohorts the perioperative and long-term outcomes of the patients were assessed. We identified 20 683 patients undergoing elective iAAA repair (76.4% EVAR). The propensity matched cohorts included 4886 pairs of patients.

Results

The perioperative mortality was 1.9% for EVAR and 5.9% for OAR (P = <.001). The perioperative mortality was mainly influenced by patients age (Odds-Ratio (OR):1.073, confidence interval (CI):1.058-1.088, P ≤ .001) and OAR (OR:3.242, CI:2.552-4.119, P ≤ .001). The early survival benefit after endovascular repair persisted for approximately 3 years (estimated survival EVAR 82.3%, OAR 80.9%, P = .021). After that time the estimated survival curves were similar. After 9 years the estimated survival was 51.2% after EVAR as compared to 52.8% after OAR (P = .102). The operation method didn’t influence long-term survival significantly (Hazard-Ratio (HR): 1.046, CI: .975-1.122, P = .211). The vascular reintervention rate was 17.4% in the EVAR cohort and 7.1% in the OAR cohort (P ≤ .001).

Conclusion

EVAR has a significantly lower perioperative mortality than OAR, a survival benefit that lasts up to 3 years after intervention. Thereafter, no significant difference in survival was observed between EVAR and OAR. The decision between EVAR or OAR may depend on patient preference, surgeons’ experience, and the institutions’ ability to handle complications.

Keywords

intact abdominal aortic aneurysm endovascular repair open repair mortality long-term survival

Introduction

The superiority of endovascular abdominal aortic aneurysm repair (EVAR) in comparison to open aneurysm repair (OAR) with respect to the 30-day-survival-rate and perioperative complications is generally accepted, however, results on long-term survival are not consistent. In EVAR trial 1 a lower mortality after OAR than after EVAR in the 8-year follow-up period was observed.¹ In contrast, in the Over trial, long-term overall survival was similar among patients who underwent endovascular repair compared to patients who underwent open surgery.² Bulder et al³ detected no difference in long-term survival between EVAR and OAR up to 10 years after surgery, whereas Li et al⁴ found a higher mortality with EVAR after 5 to 9 years. Furthermore, they found a higher reintervention rate and a higher secondary rupture rate with EVAR. Many confounding factors make it difficult to compare the 2 procedures in routine clinical practice. To make the 2 procedures more comparable, Schermerhorn et al⁵ analyzed perioperative and long-term survival, reinterventions, and complications after EVAR and OAR in Medicare beneficiaries in propensity score matched cohorts. In their study, the endovascular repair was associated with a significant early survival benefit compared with open repair, which gradually decreased over time. A significantly higher rate of late rupture was observed after endovascular repair than after open repair. Whether similar results can be observed in Germany in a patient population that is treated with EVAR in a significantly later period and with increasing experience with EVAR, will be examined in this study. Therefore, the perioperative complications and long-term survival after elective open (OAR) and endovascular (EVAR) repair of intact abdominal aortic aneurysms (iAAA) were analyzed using data from the largest German health insurance company, which serves about 1 third of the population (approx. 27 mill people).

Methods

In this retrospective study, anonymized patient data of “AOK-Die Gesundheitskasse” registry (largest statutory health insurance company in Germany) was analyzed. Since health insurance data was analyzed, the entire medical history of every patient was considered. Therefore, this study was not limited to 1 hospital group or medical department. Patients’ medical information was coded using the International Classification of Diseases, 10th Revision (ICD-10) and Operations and Procedures Codes (OPS). Data from all patients who underwent endovascular (OPS code: 5-38a.1) (n = 15 792) or open repair (OPS codes: 5-384.7) (n = 4891) for an iAAA (ICD code: I71.4 abdominal aorta aneurysm, without mention of rupture) between Jan. 1, 2010, and Dec. 31, 2016, were included in the study. Comorbidities, perioperative complications, and complications in the follow-up were analyzed using the documented ICD and OPS codes. All patients were followed until December 31, 2018. The mean follow-up time of the patient cohort was 79 months (minimum follow-up: 0 months, maximum follow-up 107 months). Patients’ death but not the cause of death was reported.

Propensity Score Matching

The propensity score matching was done by using R Software (Foundation for Statistical Computing, Vienna, Austria). The R packages “MatchIt” with the method “nearest” were used to calculate the propensity scores and logistic regression models were created to predict the likelihood of endovascular repair to ensure compatibility between the EVAR and OAR groups to minimize bias. All comorbidities analyzed in Tables 1 and 2 were included as explanatory variables. It was ensured that the included comorbidities were diagnosed before the operation. Additionally, the gender and the age of the patient were also included as variables. Furthermore, it was decided that a matching without a replacement would be best to perform so that it can be guaranteed that every EVAR-patient gets matched to 1 patient from the OAR group.

Table 1.

Baseline Characteristics of Patients Undergoing Endovascular Repair or Open Repair of Intact Abdominal Aortic Aneurysms Before Propensity Score Matching.

Characteristics and comorbidities	EVAR (n = 15.792)	OAR (n = 4891)	P-value
Male sex, n (%)	13.666 (86.5)	4080 (83.4)	<.001
Female sex, n (%)	2126 (13.5)	811 (16.6)	<.001
Age, mean±SD in years, median (min-max)	73.6 ± 8.4, 75 (38-98)	70.3 ± 8.6, 71 (16-96)	<.001
Age men, mean±SD in years, median (min-max)	73.1 ± 8.4, 74 (38-98)	69.8 ± 8.5, 71 (33-96)	<.001
Age women, mean±SD in years, median (min-max)	76.6 ± 7.9, 77 (40-95)	72.9 ± 8.7, 74 (16-91)	<.001
Patients ≥80 years, n (%)	3934 (24.9)	666 (13.6)	<.001
Patients ≥70 years, n (%)	11.330 (71.7)	2871 (58.7)	<.001
History of myocardial infraction, n (%)	2152 (13.6)	442 (9.0)	<.001
History of stroke, n (%)	752 (4.8)	185 (3.8)	.004
History of intracerebral bleeding, n (%)	57 (.4)	11 (.2)	.201
History of TIA, n (%)	397 (2.5)	95 (1.9)	.022
Arterial hypertension, n (%)	8658 (54.8)	2100 (42.9)	<.001
Dyslipoproteinemia, n (%)	5280 (33.4)	1272 (26.0)	<.001
Diabetes mellitus type 2, n (%)	2780 (17.6)	568 (11.6)	<.001
COPD, n (%)	2332 (14.8)	518 (10.6)	<.001
Renal insufficiency (stage 3-5), n (%)	1702 (10.8)	365 (7.5)	<.001
- Stage 3, n (%)	−1473 (9.3)	−302 (6.2)	<.001
- Stage 4, n (%)	−145 (.9)	−44 (.9)	.994
- Stage 5, n (%)	−84 (.5)	−19 (.4)	.247
Left heart failure (NYHA 2-4 and unspecified), n (%)	2074 (13.1)	434 (8.9)	<.001
- NYHA 2, n (%)	−629 (4.0)	−147 (3.0)	.002
- NYHA 3, n (%)	−798 (5.1)	−158 (3.2)	<.001
- NYHA 4, n (%)	−518 (3.3)	−105 (2.1)	<.001
- Unspecified, n (%)	−129 (.8)	−24 (.5)	.024
PAD (fontaine stage 3-4), n (%)	668 (4.2)	186 (3.8)	.203
- Stage 3, n (%)	−344 (2.2)	−144 (2.3)	.543
- Stage 4, n (%)	−324 (2.1)	−72 (1.5)	.011

EVAR Endovascular aneurysm repair, OAR Open aneurysm repair, SD standard deviation , min-max: minimum-maximum, TIA Transient ischemic attack, COPD chronic obstructive pulmonary disease, NYHA New York Heart Association Renal insufficiency stage 3-5 Glomerular filtration rate under 60 ml/min/1.73 m², Renal insufficiency stage 3 Glomerular filtration rate from 30 to 59 ml/min/1.73 m², Renal insufficiency stage 4 Glomerular filtration rate from 15 to 29 ml/min/1.73 m², Renal insufficiency stage 5 Glomerular filtration rate under 15 ml/min/1.73 m², PAD peripheral artery disease.

Table 2.

Baseline Characteristics of Patients Undergoing Endovascular Repair or Open Repair of Intact Abdominal Aortic Aneurysms After Propensity Score Matching.

Characteristics and Comorbidities	EVAR (n = 4886)	OAR (n = 4886)	P-value
Male sex, n (%)	4100 (83.9)	4078 (83.5)	.565
Female sex, n (%)	786 (16.1)	808 (16.5)	.565
Age, mean±SD in years, median (min-max)	70.5 ± 8.6, 72 (38-96)	70.3 ± 8.5, 71 (36-96)	.457
Age men, mean±SD in years, median (min-max)	69.9 ± 8.5, 71 (38-96)	69.8 ± 8.5, 71 (36-96)	.629
Age women, mean±SD in years, median (min-max)	73.5 ± 8.0, 74 (40-93)	73.1 ± 8.2, 74 (45-91)	.379
Patients ≥80 years, n (%)	684 (14.0)	666 (13.6)	.598
Patients ≥70 years, n (%)	2905 (59.5)	2871 (58.8)	.497
History of myocardial infraction, n (%)	487 (10.0)	442 (9.0)	.121
History of stroke, n (%)	187 (3.8)	185 (3.8)	.958
History of intracerebral bleeding, n (%)	12 (.2)	11 (.2)	.839
History of TIA, n (%)	78 (1.6)	95 (1.9)	.193
Arterial hypertension, n (%)	2067 (42.3)	2100 (43.0)	.500
Dyslipoproteinemia, n (%)	1256 (25.7)	1272 (26.0)	.729
Diabetes mellitus type 2, n (%)	538 (11.0)	568 (11.6)	.354
COPD, n (%)	472 (9.7)	518 (10.6)	.131
Renal insufficiency (stage 3-5), n (%)	320 (6.5)	365 (7.5)	.081
- Stage 3, n (%)	−266 (5.4)	−302 (6.2)	.120
- Stage 4, n (%)	−36 (.7)	−44 (.9)	.432
- Stage 5, n (%)	−18 (.4)	−19 (.4)	1.000
Left heart failure (NYHA 2-4 and unspecified), n (%)	386 (7.9)	434 (8.9)	.086
- NYHA 2, n (%)	−121 (2.5)	−147 (3.0)	.121
- NYHA 3, n (%)	−152 (3.1)	−158 (3.2)	.773
- NYHA 4, n (%)	−90 (1.8)	−105 (2.1)	.311
- Unspecified, n (%)	−23 (.5)	−24 (.5)	1.000
PAD (fontaine stage 3-4), n (%)	172 (3.5)	186 (3.8)	.452
- Stage 3, n (%)	−90 (1.8)	−114 (2.3)	.103
- Stage 4, n (%)	−82 (1.7)	−72 (1.5)	.465

EVAR Endovascular aneurysm repair, OAR Open aneurysm repair, SD standard deviation , min-max: minimum-maximum, TIA Transient ischemic attack, COPD chronic obstructive pulmonary disease, NYHA New York Heart Association, Renal insufficiency stage 3-5 Glomerular filtration rate under 60 ml/min/1.73 m², Renal insufficiency stage 3 Glomerular filtration rate from 30 to 59 ml/min/1.73 m², Renal insufficiency stage 4 Glomerular filtration rate from 15 to 29 ml/min/1.73 m², Renal insufficiency stage 5 Glomerular filtration rate under 15 ml/min/1.73 m², PAD peripheral artery disease.

Statistical Analysis

After using R for the propensity score matching the rest of the analysis was completed with SPSS 27 (IBM Deutschland GmbH, Ehningen, Deutschland). This study was designed according to the intention-to-treat principle. To analyze if there were significant differences between both operation methods Chi-square-tests were used for non-metric variables. The P-value corresponded to the significance of the Fisher’s exact test. For metric variables a Mann-Whitney-U-test was performed. Those tests were also performed to ensure that there were no significant differences between the operations methods regarding the baseline patients’ characteristic after the propensity score matching. To analyze the effect of patients’ characteristics and comorbidities on perioperative mortality a univariable binary logistic regression was performed. All statistically significant variables (P < .05) from the univariable model were included in a multivariable analysis. To estimate overall survival Kaplan-Meier tables were created. Comparisons were made using the log-rank test. We also performed separate survival analyses for 3, 5, and 9 years of follow-up. All complications and aneurysm-related reinterventions, which occurred during the follow-up period were also estimated using the Kaplan-Meier life-table method and compared with the log-rank test. Complications or reinterventions that occurred in the perioperative period were not integrated in this analysis. Univariable Cox proportional model was performed prior to a multivariable Cox proportional model to evaluate whether comorbidities, age and type of surgery had an impact on overall survival. All parameters, which had a statistically significant impact on survival in the univariable analysis (P < .05) were tested in the multivariable analysis. Log-rank tests were performed to test for significance. P-values of less than .05 were considered statistically significant.

Results

Patients

The study population consisted of a total of 20 683 patients. 15 792 (76.4%) underwent EVAR and 4891 (23.6%) OAR (Table 1). 17 746 (85.8%) were men, and 2937 (14.2%) were women. EVAR patients were significantly older with a mean age of 73.6 years (SD: 8.4) than OAR patients with a mean age of 70.3 years (SD:8.6) (P ≤ .001). EVAR patients had significantly more comorbidities. 2152 (13.6%) EVAR patients suffered myocardial infarction before surgery compared to 442 (9.0%) OAR patients (P ≤ .001). EVAR patients were also more likely to have chronic obstructive pulmonary disease (COPD) than OAR patients (EVAR: 14.8%; OAR: 10.6%; P ≤ .001)

The propensity matched cohort included 4886 pairs of patients (mean age: 70.4 years SD: 8.5). There were no significant differences in baseline covariates between EVAR and OAR cohorts (Table 2).

Perioperative Outcome

Perioperative mortality was 1.9% after endovascular and 5.9% after open repair (P ≤ .001) (Table 3). The benefit of lower mortality with the endovascular repair was seen in all analyzed age groups and in both, men and women. The length of hospital stay was significantly shorter (EVAR mean: 13.0 days, OAR mean: 20.3 days, P ≤ .001) and the rate of patients requiring blood transfusions was significantly lower with EVAR (14.0% vs 60.8%, P ≤ .001). The endovascular repaired cohort also had significantly lower rates of perioperative medical complications (eg, pneumonia 2.6% vs 10.5%, P ≤ .001, acute renal failure 2.9% vs 9.3%, P ≤ .001, myocardial infarction 1.4% vs 2.7%, P = < .001) and surgical complications (eg, mesenteric ischemia .6% vs 2.7%, P ≤ .001, postoperative ileus 1.1% vs 5.2%, P ≤ .001) compared with the OAR cohort.

Table 3.

Perioperative Outcomes After Endovascular Repair or Open Repair of Intact Abdominal Aortic Aneurysm in a Propensity Score Matched Study Population.

	EVAR (n = 4886)	OAR (n = 4886)	P-value
Perioperative mortality, n (%)	93 (1.9)	286 (5.9)	<.001
Perioperative mortality men, n (%)	74/4100 (1.8)	234/4078 (5.7)	<.001
Perioperative mortality women, n (%)	19/786 (2.4)	52/808 (6.4)	<.001
Perioperative mortality ≥80 years, n (%)	23/684 (3.4)	97/666 (14.6)	<.001
Perioperative mortality <80 years, n (%)	70/4202 (1.7)	189/4220 (4.5)	<.001
Perioperative mortality ≥70 years, n (%)	75/2905 (2.6)	224/2871 (7.8)	<.001
Perioperative mortality <70 years, n (%)	18/1981 (.9)	62/2015 (3.1)	<.001
LOS, mean±SD in days, median (min-max)	13.0 ± 11.5, 9 (1-133)	20.3 ± 15.8, 15 (0-327)	<.001
LOS men, mean±SD in days, median (min-max)	12.8 ± 11.3, 8 (1-133)	20.1 ± 15.8, 15 (0-327)	<.001
LOS women, mean±SD in days, median (min-max)	14.4 ± 12.5, 9 (1-123)	21.6 ± 15.5, 17 (1-183)	<.001
Blood transfusions, n (%)	683 (14.0)	2969 (60.8)	<.001
Intensive care treatment, n (%)	834 (17.1)	2918 (59.7)	<.001
Wound complications, n (%)	181 (3.7)	384 (7.9)	<.001
Myocardial infarction, n (%)	70 (1.4)	132 (2.7)	<.001
Cerebral complications, n (%)	54 (1.1)	73 (1.5)	.108
Pneumonia, n (%)	128 (2.6)	511 (10.5)	<.001
Acute renal failure, n (%)	143 (2.9)	453 (9.3)	<.001
Dialysis, n (%)	83 (1.7)	283 (5.8)	<.001
Deep-vein thrombosis, n (%)	23 (.5)	30 (.6)	.409
Arterial embolism and thrombosis of the lower limbs, n (%)	111 (2.3)	236 (4.8)	<.001
Tracheostomy, n (%)	24 (.5)	132 (2.7)	<.001
Mesenteric ischemia, n (%)	31 (.6)	131 (2.7)	<.001
Major amputation, n (%)	5 (.1)	24 (.5)	<.001
Foot amputation, n (%)	7 (.1)	21 (.4)	.012
Bowel resection, n (%)	31 (.6)	113 (2.3)	<.001
Ileus, n (%)	56 (1.1)	253 (5.2)	<.001
Arterial embolectomy of the lower limb, n (%)	166 (3.4)	240 (4.9)	<.001

EVAR Endovascular aneurysm repair, OAR Open aneurysm repair, SD standard deviation, min-max: minimum-maximum, LOS Length of sta

Long-Term Survival

The long-term survival in the 2 cohorts is shown in Figure 1. The early survival benefit after endovascular repair persisted for approximately 3 years (EVAR estimated survival rate 82.3%, OAR 80.9%, P = .021). After that time the estimated survival curves were similar. 71.5% of EVAR patients survived 5 years compared to 70.4% of OAR patients (P = .060). After 9 years estimated survival was 51.2% among patients in the EVAR cohort as compared to 52.8% in the OAR cohort (P = .102). The survival rate depended on patients’ age as demonstrated in Table 4. There was no difference in estimated survival in patients with EVAR and OAR in the analyzed age groups, apart from patients 80 years and older. These patients showed a significantly better estimated survival after EVAR as compared to OAR (27.1% vs 25.7%, P = .048).

Figure 1.

Survival after endovascular repair or open repair of abdominal aortic aneurysm.

Table 4.

Nine-Year Outcomes After Endovascular or Open Repair of Intact Abdominal Aortic Aneurysm.

	EVAR (n = 4793)	OAR (n = 4600)	P-value
Survival all patients, n (%)	3400/4886 (51.2)	3201/4886 (52.8)	.102
Survival, patients <80 years, n (%)	3067/4202 (54.9)	2922/4220 (56.9)	.228
Survival, patients ≥80 years, n (%)	333/684 (27.1)	279/666 (25.7)	.048
Survival, patients <70 years, n (%)	1628/1981 (70.0)	1582/2015 (67.5)	.089
Survival, patients ≥70 years, n (%)	1772/2905 (39.4)	1619/2871 (43.2)	.311
Hospital readmission rate per year, mean±SD, median (min-max)	1.5 ± 1.8, 1.0 (0-24)	1.2 ± 1.7, .8 (0-18)	<.001
LOS per hospital readmission in days, mean±SD in days, median (min-max)	6.7 ± 6.9, 5 (0-103)	7.1 ± 7.8, 6 (0-135)	.037
Myocardial infarction, n (%)	429 (16.4)	420 (15.2)	.214
Cerebral complications, n (%)	535 (19.6)	524 (19.8)	.158
Dialysis, n (%)	209 (7.2)	233 (8.1)	.522
Deep-vein thrombosis, n (%)	125 (5.1)	119 (4.4)	.364
Arterial embolism and thrombosis of the lower limb n (%)	274 (9.3)	155 (5.2)	<.001
Inguinal hernia repair, n (%)	158 (6.4)	163 (5.1)	.947
Abdominal wall hernia repair, n (%)	81 (4.6)	523 (14.3)	<.001
Bowel resection, n (%)	129 (4.3)	158 (5.4)	.162
Ileus, n (%)	305 (11.5)	396 (14.2)	.005
Shunt or bypass surgery, A. Iliaca, n (%)	62 (2.1)	39 (1.3)	.010
Shunt or bypass surgery, A. femoralis, n (%)	231 (7.4)	206 (6.4)	.123
Resection or interposition, infrarenal abdominal aorta, n (%)	67 (3.2)	4 (.3)	<.001
Endovascular stent implantation, pelvic arteries, n (%)	139 (5.6)	32 (1.4)	<.001
Endovascular stent implantation, abdominal aorta, n (%)	82 (4.0)	47 (1.9)	<.001

EVAR Endovascular aneurysm repair, OAR Open aneurysm repair, SD standard deviation, min-max: minimum-maximum, LOS Length of stay.

Percentages are Kaplan-Meier estimates.

Other Long-Term Outcomes

Hospital readmission rates per year and long-term complication rates in the follow-up period are given in Table 4. Yearly hospital readmissions were more frequently seen with EVAR as compared to OAR (1.5 vs 1.2, P ≤ .001). The length of stay per readmission, however, was longer in the OAR cohort (7.1 days vs 6.7 days, P = .037). Embolism and thrombosis of the lower limb were more often seen in the EVAR cohort (9.3% vs 5.2%, P ≤ .001). In the follow-up, abdominal wall hernia repairs were performed in only 4.6% of EVAR patients but in 14.3% OAR patients (P ≤ .001), ileus occurred in 11.5% of EVAR patients vs 14.2% of OAR patients (P = .005)

The vascular reintervention rate is shown in Figure 2. 17.4% of EVAR patients underwent at least 1 additional vascular reintervention at the end of follow-up compared to only 7.1% of OAR patients (P ≤ .001).

Figure 2.

Cumulative incidence of vascular reintervention after endovascular repair or open repair of abdominal aortic aneury.

Sensitivity Analysis

Table 5 shows possible confounding factors on perioperative mortality as found in the multivariable binary logistic regression analysis. Patient age (OR: 1.073, CI: 1.058-1.088, P ≤ .001) and OAR (OR: 3.242, CI: 2.552-4.119, P ≤ .001) had a significant impact on perioperative mortality. In addition, perioperative mortality was significantly influenced by a history of transient ischemic attack (TIA) (OR: 2.502, CI: 1.484-4.219, P = .001) and renal failure stages 3-5 (OR: 1.673, CI: 1.212-2.309, P = .002).

Table 5.

Multivariate Binary Logistic Regression Analysis Regarding Perioperativ Mortality.

	Odds Ratio	95% CI	P-value
Age (increase by 1 year)	1.073	1.058-1.088	<.001
OAR (vs. EVAR)	3.242	2.552-4.119	<.001
History of TIA	2.502	1.484-4.219	.001
Arterial hypertension	1.125	.891-1.420	.321
Diabetes mellitus type 2	1.205	.887-1.637	.233
COPD	1.092	.793-1.504	.589
Renal insufficiency (stage 3-5)	1.673	1.212-2.309	.002
Left heart failure (NYHA 2-4 and unspecified)	1.305	.936-1.818	.116
PAD (fontaine stage 3-4)	1.300	.806-2.096	.282

EVAR Endovascular aneurysm repair, OAR Open aneurysm repair, CI confidence interval TIA Transient ischemic attack, COPD chronic obstructive pulmonary disease Renal insufficiency stage 3-5 Glomerular filtration rate under 60 ml/min/1.73 m², NYHA New York Heart Association, PAD peripheral artery disease.

Patient gender was not included in the multivariable binary logistic regression analysis because it had no significant influence on perioperative mortality in the univariable analysis (female gender: OR: 1.191, CI: .915-1.551, P = .194).

Factors influencing the long-term survival of the patients in the multivariable Cox-regression analysis are given in Table 6. Patients’ age (HR: 1.065, CI:1.060-1.070, P ≤ .001), history of stroke (HR: 1.239, CI:1.034-1.485, P = .020), arterial hypertension (HR: 1.180, CI:1.083-1.287, P ≤ .001), COPD (HR: 1.674, CI:1.5108-1.858, P ≤ .001), renal insufficiency stage 3-5 (HR: 1.517, CI:1.345-1.712, P ≤ .001), left heart failure (HR:1.290, CI:1.140-1.461, P ≤ .001) and peripheral artery disease (PAD) (HR:1.638, CI:1.393-1.927, P ≤ .001) had a negative influence on survival. On the other hand, dyslipoproteinemia had a positive influence on survival (HR:0.821, CI:0.744-.905, P ≤ .001). Neither gender nor type of surgical procedure influenced survival significantly.

Table 6.

Hazard Ratio and Proportional Hazard Model (Multivariate Analysis) for Overall Survival.

	HR	95% CI	P-value
Age (increase by 1 year)	1.065	1.060-1.070	<.001
Female gender (vs. male gender)	.975	.890-1.068	.579
OAR (vs. EVAR)	1.046	.975-1.122	.211
History of myocardial infarction	.996	.877-1.132	.954
History of stroke	1.239	1.034-1.485	.020
History of TIA	1.194	.920-1.550	.182
Arterial hypertension	1.180	1.083-1.287	<.001
Dyslipoproteinemia	.821	.744-.905	<.001
Diabetes mellitus type 2	1.087	.974-1.214	.136
COPD	1.674	1.508-1.858	<.001
Renal insufficiency (stage 3-5)	1.517	1.345-1.712	<.001
Left heart failure (NYHA 2-4 and unspecified)	1.290	1.140-1.461	<.001
PAD (fontaine stage 3-4)	1.638	1.393-1.927	<.001

EVAR Endovascular aneurysm repair, OAR Open aneurysm repair, HR Hazard ratio, CI confidence interval TIA Transient ischemic attack, COPD chronic obstructive pulmonary disease, Renal insufficiency stage 3-5 Glomerular filtration rate under 60 ml/min/1.73 m², NYHA New York Heart Association, PAD peripheral artery disease.

Discussion

In this propensity score matched cohort study, perioperative mortality was significantly lower after EVAR as compared to OAR, which resulted in a survival benefit that persisted up to 3 years after surgery. Subsequently, EVAR and OAR did not differ in their survival up to 9 years after surgery. This confirmed the results of the meta-analysis by Bulder et al,³ who found no difference between EVAR and OAR in long-term survival up to 10 years postoperatively. Schermerhorn et al⁵ also found in their matched cohort study only an early survival benefit up to 3 years with endovascular repair compared to open repair. After that time the survival was similar in the 2 groups. These results are also consistent with a retrospective population-based cohort study from Ontario, Canada by Salata et al.⁶ They detected 4010 propensity score matched patient pairs among 17 683 patients, who underwent elective AAA surgery. Compared with OAR, EVAR was associated with a higher survival rate up to 1 year after repair and a higher major adverse cardiovascular event–free survival rate up to 4 years after repair. Survival analyses demonstrated no statistically significant differences in long-term survival, reintervention, and secondary rupture for patients who underwent EVAR compared with those who underwent OAR. In this study, endovascular aortic repair was not associated with a difference in long-term survival even during more than 13 years’ maximum follow-up.

The present study is so far the largest cohort study with the longest follow-up reported from Germany after EVAR and OAR. German health insurance data (DAK Gesundheit) was also presented by Behrendt et al⁷ They analyzed long-term survival after EVAR (n = 3493) and OAR (n = 1457) in a smaller unmatched cohort. In their study, hospital mortality was 1.2% after EVAR and 5.4% after OAR. Kaplan-Meier analyses of survival after iAAA repair showed that the early benefit in cumulative survival after elective EVAR was reversed at about 1.5 years of follow-up. In the adjusted analyses using Cox proportional hazard models, survival was similar between the EVAR and OAR groups (HR:1.028; CI:0.889- 1.188, P = .708). This data is consistent with the results found in the matched population here (hospital mortality EVAR 1.9%, OAR 5.9%).

Varkevisser et al⁸ compared the long-term mortality between EVAR and open abdominal aneurysm repair for patients of different ages. In their study, EVAR in patients younger than 65 was associated with higher long-term all-cause mortality compared with open repair. The overall mortality was balanced between EVAR and open repair in patients aged between 65 and 79 years, and lower after EVAR as compared to OAR in patients aged 80 years. They concluded that a young, healthy patient with low operative risk and a long-life expectancy may be considered for open AAA repair, whereas EVAR remains the preferred treatment for most older AAA patients. Our results are consistent with these findings, patients younger than 70 years (EVAR: 70.0%, OAR: 67.5%, P = .089) and older than 70 years (EVAR: 39.4%, OAR: 43.2%, P = .311) as well as patients younger than 80 years (EVAR: 54.9%, OAR: 56.9%, P = .228) did not significantly differ in their long-term survival with respect to the surgical procedure (EVAR or OAR). Only patients, who were 80 years and older, survived significantly longer after EVAR than after OAR (EVAR: 27.1%, OAR: 25.7% P = .048). Based on that our results favor EVAR for patients of older age. Especially given the high hospital mortality of OAR, the implication must be to offer OAR only in high-volume centers with low hospital mortality, at least for older and high-risk patients.⁹ Anyhow, the 5.9% OAR hospital mortality observed here does not meet the standard required by the SVS guidelines,¹⁰ which calls for perioperative mortality of 5% or less.

The long-term survival and cardiovascular morbidity between women and men undergoing elective abdominal aortic aneurysm (AAA) repair at National Health Service hospitals in England was reported by Desai et al¹¹ Female patients had higher 30-day mortality with both open repair (OR 1.39, CI 1.25-1.56, P < .001) and EVAR (OR 1.57, CI 1.23-2.00, P < .001) compared to men. Women patients had higher 30-day, 1-year, 5-year, and aortic-related mortalities compared with men for both types of AAA repair. O`Donnell et al¹² compared long-term survival, aneurysm-related mortality, and rates of endoleaks and reinterventions between male and female patients in the Endurant Stent Graft Natural Selection Global Postmarket Registry (ENGAGE) using 2:1 propensity score matching. In their study, female patients had long-term outcomes comparable to those of male patients. In our study female sex was not a risk factor, neither for perioperative mortality (univariable analyses female sex: OR: 1.191, CI: .915-1.551, P = .194) nor for long-term outcome (female sex: HR:0.975, CI:0.890-1.068, P = .579). Behrendt et al¹³ also concluded that short- or long-term survival and reintervention outcomes were similar in both sexes.

Schermerhorn et al⁵ noted a significantly increased risk of rupture with EVAR (5.4%) compared to OAR (1.4%). Similar results are reported by Yei et al.¹⁴ In their analysis of the Society for Vascular Surgery (SVS) Vascular Quality Initiative (VQI) database, regarding a total of 32 760 patients, 2852 patients each with EVAR and OAR were included in a propensity score matched comparison. In their study, endovascular repair was associated with significantly higher rates of long-term rupture and reintervention. We were not able to comment on the long-term aneurysm rupture rate with endovascular and open repair, because the causes of death of the patients were not reported. Nevertheless, in our study the hospital readmission rate per year was also significantly higher after EVAR with 1.5 ± 1.8 than after OAR with 1.2 ± 1.7 (P ≤ .001). In contrast, the hospital stay per readmission was significantly shorter for EVAR (6.7 ± 6.9 days vs 7.1 ± 7.8 days, P = .037). Additionally, we determined vascular reinterventions, they were significantly (P ≤ .001) higher in EVAR patients (17.4%) as compared to OAR patients (7.1%) (Figure 2). However, other reasons for readmission were less frequently seen with EVAR as compared to OAR (ileus: 11.5% vs 14.2%, P = .005; abdominal wall hernia repair 4.6% vs 14.3%, P ≤ .001).

Limitations of this study include potential coding errors. The cause of death was not reported in the given dataset, therefore aneurysm-related deaths could not be specified. The endoleak-rate and aneurysm-rupture rate were not documented.

Furthermore, hospital mortality in patients over 80 years of age seems to be very high with OAR (14.6%). Whether the AOK data represent the overall standard quality of open aortic aneurysm repair in Germany is unknown. The levels of the treating hospitals and the percentage of patients treated in high- and low-volume hospitals were unknown.

Information on aneurysm anatomy (juxtarenal/infrarenal) and aneurysm size is not reported in this study

Conclusion

Summarizing these results, the in-hospital mortality was significantly lower after EVAR than after OAR, which resulted in a survival benefit that persisted up to 3 years after surgery. Nevertheless, at a long-term follow-up of 9 years, there were no significant differences between OAR and EVAR in terms of patient survival. Thus, we suggest that the decision between EVAR and OAR should depend on the patient’s preference and given the high hospital mortality of OAR in elderly patients, on the surgeon’s experience.

Footnotes

Author Contributions

All authors contributed to the study conception and design. Material preparation and data collection were performed by Jasmin Epple and Reinhart T. Grundmann. The statistical analysis was performed by Jasmin Epple and Neelam Lingwal. The first draft of the manuscript was written by Reinhart T. Grundmann and Jasmin Epple, involving all authors. The final version of this paper was approved by all authors.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Jasmin Epple

References

Patel

Sweeting

Powell

Greenhalgh

. Endovascular versus open repair of abdominal aortic aneurysm in 15-years’ follow-up of the UK endovascular aneurysm repair trial 1 (EVAR trial 1): A randomised controlled trial. Lancet. 2016;388(10058):2366-2374.

Lederle

Kyriakides

Stroupe

, et al. Open versus endovascular repair of abdominal aortic aneurysm. N Engl J Med. 2019;380:2126-2135.

Bulder

RMA

Bastiaannet

Hamming

Lindeman

JHN

. Meta-analysis of long-term survival after elective endovascular or open repair of abdominal aortic aneurysm. Br J Surg. 2019;106:523-533.

Khan

Salata

, et al. A systematic review and meta-analysis of the long-term outcomes of endovascular versus open repair of abdominal aortic aneurysm. J Vasc Surg. 2019;70:954-969.e30.

Schermerhorn

Buck

O’Malley

, et al. Long-term outcomes of abdominal aortic aneurysm in the medicare population. N Engl J Med. 2015;373:328-338.

Salata

Hussain

de Mestral

, et al. Comparison of outcomes in elective endovascular aortic repair vs open surgical repair of abdominal aortic aneurysms. JAMA Netw Open. 2019;2:e196578.

Behrendt

Sedrakyan

Rieß

, et al. Short-term and long-term results of endovascular and open repair of abdominal aortic aneurysms in Germany. J Vasc Surg. 2017;66:1704-1711.

Varkevisser

RRB

Carvalho Mota

Swerdlow

, et al. Long-term age-stratified survival following endovascular and open abdominal aortic aneurysm repair. J Vasc Surg. 76, 2022;899-907.

Trenner

Kuehnl

Salvermoser

, et al. Editor’s choice - high annual hospital volume is associated with decreased in hospital mortality and complication rates following treatment of abdominal aortic aneurysms: Secondary data analysis of the nationwide german DRG Statistics from 2005 to 2013. Eur J Vasc Endovasc Surg. 2018;55:185-194.

10.

Chaikof

Dalman

Eskandari

, et al. The society for vascular surgery practice guidelines on the care of patients with an abdominal aortic aneurysm. J Vasc Surg. 2018;67:2-77.

11.

Desai

Choke

Sayers

Nath

Bown

. Sex-related trends in mortality after elective abdominal aortic aneurysm surgery between 2002 and 2013 at national health service hospitals in England: Less benefit for women compared with men. Eur Heart J. 2016;37:3452-3460.

12.

O’Donnell

TFX

Verhagen

Pratesi

, et al. Female sex is associated with comparable 5-year outcomes after contemporary endovascular aneurysm repair despite more challenging anatomy. J Vasc Surg. 2020;71:1179-1189.

13.

Behrendt

Kreutzburg

Kuchenbecker

, et al. Female sex and outcomes after endovascular aneurysm repair for abdominal aortic aneurysm: A propensity score matched cohort analysis. J Clin Med. 2021;10:162.

14.

Yei

Mathlouthi

Naazie

Elsayed

Clary

Malas

. Long-term outcomes associated with open vs endovascular abdominal aortic aneurysm repair in a medicare-matched database. JAMA Netw Open. 2022;5:e2212081.