Outcomes of Endovascular Treatment for Infective Aortic Aneurysms　― A Multicenter Retrospective Study ―

Abstract

Background: In Taiwan, infective native aortic aneurysms (INAAs) are relatively common, so the aim of present study was to demonstrate the comparative outcomes of endovascular repair for thoracic and abdominal INAAs.

Methods and Results: Patients with naïve thoracic or abdominal INAAs managed with endovascular repair between 2001 and 2018 were included in this multicenter retrospective cohort. The confounding factors were adjusted with propensity score (PS). Of the 39 thoracic and 43 abdominal INAA cases, 41 (50%) presented with aneurysmal rupture, most of which were at the infrarenal abdominal (n=35, 42.7%) or descending thoracic aorta (n=25, 30.5%). Salmonella spp. was the most frequently isolated pathogen. The overall in-hospital mortality rate was 18.3%. The risks of in-hospital death and death due to rupture were significantly lower with thoracic INAAs (12.8% vs. 23.3%; PS-adjusted odds ratio (OR) 0.24, 95% confidence interval (CI) 0.06–0.96; 0.1% vs. 9.3%; PS-adjusted OR 0.11, 95% CI 0.01–0.90). During a mean follow-up of 2.5 years, the risk of all-cause death was significantly higher with thoracic INAAs (35.3% vs. 15.2%; PS-adjusted HR 6.90, 95% CI 1.69–28.19). Chronic kidney disease (CKD) was associated with death.

Conclusions: Compared with thoracic INAAs, endovascular repair of abdominal INAAs was associated with a significantly higher in-hospital mortality rate. However, long-term outcomes were worse for thoracic INAAs, with CKD and infections being the most important predictor and cause of death, respectively.

Infective native aortic aneurysms (INAAs, previously called mycotic aortic aneurysms [AAs]) are known for their high mortality rate and severe long-term infection-related sequalae. First described in 1885 by William Osler, INAAs remain a clinical challenge in modern medicine. In addition to being rare (0.7–4.5% of all AAs),¹ the heterogeneity of causative microbes and anatomic locations further preclude the establishment of standardized practice guidelines. Tailored therapy is imperative because patients with INAAs are often immunocompromised and frail with multiple comorbidities. In Taiwan, INAA is relatively common, with prevalence up to 13.3% of all AAs in 1 report.²

Endovascular techniques have largely supplanted open repair for degenerative AAs, and their adoption for the treatment of INAAs has increased over the past 20 years.^3–10 Although this relatively well-tolerated approach provides more aggressive management for patients too frail to undergo open repair, it raises concerns regarding infection control by leaving the infected tissue unattended, mandating lifelong antibiotic therapy in many cases. A recent systematic review concluded that short-term survival was better with endovascular aortic repair (EVAR) than open repair for abdominal INAAs, and the rates of infection-related complications were comparable.⁷ Nonetheless, subsequent studies have demonstrated conflicting results regarding the persistence of infection and its effect on late mortality and morbidity.^3–5,11 The most recent guidelines of the European Society for Vascular Surgery (ESVS) recommended EVAR with long-term postoperative antibiotics as an acceptable alternative to open repair for abdominal INAAs.¹² The American Heart Association, on the other hand, preserves the use of EVAR as a bridging or palliative measure.¹ The role of thoracic EVAR (TEVAR) for thoracic INAAs is less clear.¹⁰ Therefore, we sought to shed light on patient selection and perioperative care for endovascular repair by demonstrating the short- and long-term outcomes for both thoracic and abdominal INAAs. As a major medical institution with several country-wide hospitals, our data represent the recent status of INAAs in Taiwan.

Methods

Data Source

This retrospective cohort utilized standardized patient-level clinical data from the Chang Gung Research Database derived from the electronic medical records of 7 hospitals (including 2 medical centers) affiliated to the Chang Gung Medical Foundation (CGMF). Diagnoses were coded using the International Classification of Diseases (ICD), 9th and 10th revisions, Clinical Modification (CM) codes, respectively. The affiliated hospitals of the CGMF account for 21.2% and 12.4% of the nation’s outpatient and inpatient visits, respectively.¹³

Study Population

Patients with a first-time diagnosis of thoracic or abdominal INAA who underwent endovascular repair at any hospital of the CGMF between November 1, 2001, and November 30, 2018, were consecutively included in this study. No ICD diagnostic codes exist for INAAs. Therefore, 2 authors (C.-C.L. and S.-W.C.) verified the diagnosis of INAA and the execution of surgical procedures by examining the inpatient medical records and operation reports of all patients admitted for AAs. INAA was defined in accordance with the ESVS guidelines,^12,14 and verified with consistent intraoperative findings. Specifically, a “probable” diagnosis of INAA was defined as meeting 2 of the following: clinical symptoms and signs (fever, abdominal or back pain, sepsis, shock), laboratory studies (elevated inflammatory markers, leukocytosis, positive blood cultures), and computer tomography (CT) characteristics (saccular, multilobular or eccentric aneurysm, peri-aortic gas or soft tissue mass, rapid expansion and/or rupture, atypical location or multiple aneurysms). “Definitive” diagnosis of INAA was defined as a “probable” INAA and consistent intraoperative findings. Those who received conservative treatment, open repair, or combined TEVAR/EVAR were excluded. By definition, all patients who underwent surgical intervention (TEVAR, EVAR or open repair) were “definitive” cases of INAA, and those managed with medical treatment alone had a diagnosis of “probable” INAA.

Covariates

Covariates were demographics (age, sex, smoking, alcohol use, and body mass index), comorbidities (listed in Table 1), clinical manifestations (pain, fever/chills, shock), laboratory results (white blood cell [WBC] count, C-reactive protein [CRP], blood culture), and the use of intravenous (IV) and oral antibiotics. Surgical details were collected, including stent types and brands, proximal and distal landing zones, adjunctive procedures (e.g., chimney, bypass, octopus), and total radiation dose and contrast volume.

Table 1. Baseline Characteristics of the Study Patients

Variable	Total (n=82)	TEVAR (n=39)	EVAR (n=43)	P value
Age, years	69.5±11.9	69.5±11.6	69.5±12.2	0.997
Male	61 (74.4)	25 (64.1)	36 (83.7)	0.042
Smoking	38 (46.3)	13 (33.3)	25 (58.1)	0.024
Alcohol	17 (20.7)	7 (17.9)	10 (23.3)	0.554
BMI, kg/m2 (n=74)	23.8 [21.1, 26.7]	24.7 [21.1, 27.1]	23.6 [21.1, 26.2]	0.555
Time of surgery				0.082
2006–2010	14 (17.1)	6 (15.4)	8 (18.6)
2011–2014	32 (39.0)	11 (28.2)	21 (48.8)
2015–2018	36 (43.9)	22 (56.4)	14 (32.6)
Comorbidity
Hypertension	56 (68.3)	26 (66.7)	30 (69.8)	0.763
Coronary artery disease	18 (22.0)	9 (23.1)	9 (20.9)	0.815
Dyslipidemia	24 (29.3)	12 (30.8)	12 (27.9)	0.776
Diabetes mellitus	34 (41.5)	17 (43.6)	17 (39.5)	0.710
Chronic kidney disease	32 (39.0)	14 (35.9)	18 (41.9)	0.580
Dialysis	8 (9.8)	4 (10.3)	4 (9.3)	0.884
Liver disease	17 (20.7)	10 (25.6)	7 (16.3)	0.296
Atrial fibrillation	7 (8.5)	4 (10.3)	3 (7.0)	0.596
Chronic obstructive pulmonary disease	11 (13.4)	2 (5.1)	9 (20.9)	0.036
Stroke	8 (9.8)	5 (12.8)	3 (7.0)	0.373
Autoimmune disease	4 (4.9)	2 (5.1)	2 (4.7)	0.920
Malignancy	6 (7.3)	4 (10.3)	2 (4.7)	0.330
Previous aortic surgery	1 (1.2)	0 (0.0)	1 (2.3)	0.338
Use of steroids	7 (8.5)	1 (2.6)	6 (14.0)	0.065
Presenting symptoms
Fever/chills	44 (53.7)	21 (53.8)	23 (53.5)	0.974
Hypotension/shock	13 (15.9)	5 (12.8)	8 (18.6)	0.474
Pain related to aneurysm
Chest pain	10 (12.2)	9 (23.1)	1 (2.3)	0.004
Abdominal pain	38 (46.3)	9 (23.1)	29 (67.4)	<0.001
Back pain	26 (31.7)	8 (20.5)	18 (41.9)	0.038
Dyspnea	11 (13.4)	7 (17.9)	4 (9.3)	0.251
Hemoptysis	5 (6.1)	5 (12.8)	0 (0.0)	0.015
Concomitant infection
Respiratory	9 (11.0)	5 (12.8)	4 (9.3)	0.611
Genitourinary	9 (11.0)	6 (15.4)	3 (7.0)	0.224
Soft tissue	13 (15.9)	6 (15.4)	7 (16.3)	0.912
Blood culture performed				0.577
No	29 (35.4)	15 (38.5)	14 (32.6)
Yes	53 (64.6)	24 (61.5)	29 (67.4)
Blood culture typing (n=53)				0.612
Salmonella sp.	12 (22.6)	4 (16.7)	8 (27.6)
Non-Salmonella	9 (17.0)	4 (16.7)	5 (17.2)
Culture-negative	32 (60.4)	16 (66.7)	16 (55.2)
Follow-up, years	0.7 [0.0, 3.4]	0.6 [0.1, 2.4]	0.7 [0.0, 5.1]	0.837

Data are presented as frequency (percentage), mean±standard deviation or median [25^th, 75^th percentile]. BMI, body mass index; EVAR, endovascular aortic repair; IV, intravenous; TEVAR, thoracic endovascular aortic repair.

Outcomes

In-hospital outcomes included in-hospital (all-cause) death, death due to rupture, reoperation, new-onset stroke, new-onset dialysis, ventilation duration, length of intensive care unit stay, and length of hospital stay. Follow-up outcomes include all-cause death, death due to rupture, infection-related death, reoperation, and readmission for aortic disease. Infection-related deaths included any infections with a significant contribution to death, as validated retrospectively through review of medical records. These include unresolved aortic infection (excluding those with eventual aneurysmal rupture) and sepsis from any source such as severe pneumonia. Patients were followed up from the day of discharge to December 31, 2018 or the day of event occurrence.

Statistical Analysis

Baseline characteristics, CT reports, laboratory results, and surgical details of patients with thoracic and abdominal INAAs were compared using Fisher’s exact test for categorical variables, independent sample t-test for continuous variables, and the Mann-Whitney U test for obviously skewed continuous variables. In-hospital outcomes were compared using logistic regression for categorical outcomes or linear regression for continuous outcomes. The risk of follow-up outcomes was compared using a Cox proportional hazards model. To control for potential confounding factors, the propensity score (PS) was calculated by regressing the study groups on baseline demographics (age, sex, smoking, and alcohol use), comorbidities, rupture and shock. Surgical year (2006–2010, 2011–2014, 2015–2018) was included in the PS calculation for trend analysis. A PS-adjusted multivariable analysis was performed for regression. Furthermore, we compared the risk of in-hospital and late outcomes for patients who underwent different treatments, including open repair, endovascular stent and medical alone. Because the sample size was very small, no covariate adjustment was done in the analysis. Finally, the potential risk factors of in-hospital death and total death (including in-hospital death and death after surviving to discharge) were explored using logistic regression and Cox model, respectively. A series of univariate analysis were conducted and variables with significance <0.2 were further included in the multivariable analysis with backward elimination procedure. A two-sided P value <0.05 was considered statistically significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

Results

Epidemiology

A total of 2,494 patients with AAs were identified, of which INAAs constituted 5.5% (n=136). Among them, 11.7% (n=42) of all endovascularly treated thoracic AAs (n=359) and 6.3% (n=46) of all endovascularly treated abdominal AAs (n=731) were INAAs. Patients managed with conservative therapy (n=22), open repair (n=29), or combined TEVAR/EVAR (n=3) were excluded. Thus, 82 cases, including 39 cases of thoracic INAAs and 43 cases of abdominal INAAs, were included in the final analysis (Figure 1).

Figure 1.

Study flowchart (A) and treatments of infective native aortic aneurysms during the study period (B). EVAR, endovascular aortic repair; TEVAR, thoracic endovascular aortic repair.

Patients’ Characteristics

Baseline characteristics are listed in Table 1, Supplementary Table 1 and Supplementary Table 2. The mean age at presentation for both groups was 69.5 years. Male predominance was more pronounced in the abdominal group (83.7% vs. 64.1%). The major comorbidities were hypertension (n=56, 68.3%) and immunocompromised status, including diabetes mellitus (n=34, 41.5%), chronic kidney disease (CKD) (n=32, 39.0%), chronic liver disease (n=17, 0.7%), malignancy (n=6, 7.3%), and glucocorticoid use (n=7, 8.5%). Psoas abscesses were more prevalent with abdominal INAAs (11.6% vs. 2.6%). Salmonella sp. was the most frequently isolated pathogen (n=12), followed by Staphylococcus sp. (n=2), gram-positive rods (n=2) and Escherichia coli (n=2) (Supplementary Table 1). IV 3rd-generation cephalosporins (n=74, 90.2%) and anti-methicillin-resistant Staphylococcus aureus (MRSA) agents (n=39, 47.6%) were the most prescribed. Use of fluoroquinolones was common in both the inpatient (n=26, 31.7%) and outpatient (n=22, 32.8%) setting (Supplementary Table 2).

Initial Imaging and Laboratory Results

Imaging and laboratory results at presentation are shown in Table 2 and Supplementary Table 3. Aneurysmal rupture was common (n=41, 50.0%). Most aneurysms were located at the infrarenal abdominal (n=35, 42.7%) or descending thoracic aorta (n=25, 30.5%); 7 patients had juxtarenal aneurysms and 15 involved ascending aorta and arch. Saccular/eccentric aneurysms were more common in abdominal INAAs (n=9), whereas thoracic INAAs were associated with more periaortic gas (n=5). WBC counts and CRP levels were significantly elevated in both groups.

Table 2. Imaging and Procedural Details

Variable	Total (n=82)	TEVAR (n=39)	EVAR (n=43)	P value
Ruptured	41 (50.0)	23 (59.0)	18 (41.9)	0.122
Multiple	4 (4.9)	3 (7.7)	1 (2.3)	0.260
Diameter, cm				0.871
≥5	26 (31.7)	12 (30.8)	14 (32.6)
<5	23 (28.0)	12 (30.8)	11 (25.6)
Unknown	33 (40.2)	15 (38.5)	18 (41.9)
Mean size diameter, cm (n=49)	5.2±2.3	5.3±2.4	5.2±2.2	0.814
Location of aneurysm
Ascending aorta/aortic arch	15 (18.3)	15 (38.5)	0 (0.0)	<0.001
Descending aorta	25 (30.5)	25 (64.1)	0 (0.0)	<0.001
Thoracoabdominal aorta	4 (4.9)	3 (7.7)	1 (2.3)	0.260
Infrarenal abdominal aorta	35 (42.7)	0 (0.0)	35 (81.4)	<0.001
Common iliac artery	3 (3.7)	0 (0.0)	3 (7.0)	0.093
Juxtarenal	6 (7.3)	0 (0.0)	6 (14.0)	0.015
Morphology
Saccular/eccentric	13 (15.9)	4 (10.3)	9 (20.9)	0.186
Lobulated/multilobular	6 (7.3)	3 (7.7)	3 (7.0)	0.901
Soft tissue mass/abscess	31 (37.8)	16 (41.0)	15 (34.9)	0.567
Periaortic gas	6 (7.3)	5 (12.8)	1 (2.3)	0.068
Fistula formation	1 (1.2)	0 (0.0)	1 (2.3)	0.338
Psoas abscesses	6 (7.3)	1 (2.6)	5 (11.6)	0.116
Procedural details
Stent type				0.012
Cook/Zenith	25 (30.5)	16 (41.0)	9 (20.9)
Medtronic	17 (20.7)	3 (7.7)	14 (32.6)
Gore	40 (48.8)	20 (51.3)	20 (46.5)
TEVAR: proximal landing (n=39)				–
Zone 0	–	4 (10.3)	–
Zone 1	–	5 (12.8)	–
Zone 2	–	10 (25.6)	–
Zone 3	–	20 (51.3)	–
TEVAR: adjunctive procedure (n=39)
Chimney	–	11 (28.2)	–	–
Bypass	–	4 (10.3)	–	–
Revascularization	–	0 (0.0)	–	–
EVAR: adjunctive procedure (n=43)
Octopus	–	–	2 (4.7)	–
Chimney	–	–	1 (2.3)	–
Total radiation dose (n=15)	1,586 [1,012, 3,641]	1,838 [1,189, 4,169]	698 [295, 2,134]	0.078
Total contrast volume (n=16)	48 [30, 65]	50 [30, 70]	30 [30, 50]	0.441

Data are presented as frequency (percentage), mean±standard deviation or median [25^th, 75^th percentile]. ALT, alanine aminotransferase; AST, aspartate aminotransferase; CRP, C-reactive protein; EVAR, endovascular aortic repair; TEVAR, thoracic endovascular aortic repair; WBC, white blood cell count.

Surgical Details

Procedural details are summarized in Table 2. In the thoracic group, proximal landing at zones 0–2 constituted half of the cases (48.7%, n=19). Adjunctive chimney procedure or bypass was performed in 11 (28.2%) and 4 (10.3%) cases of thoracic INAAs, respectively. Octopus and chimney procedures were performed in 2 (4.7%) and 1 (2.3%) case(s) of abdominal INAAs, respectively.

In-Hospital Outcomes

Duration of preoperative antibiotic was 34.5 (standard deviation [SD]=19.6) and 32.5 (SD=24.2) days in the thoracic and abdominal groups, respectively. Table 3 presents the in-hospital outcomes. PS-adjusted analysis demonstrated significantly lower risk of in-hospital death with thoracic INAAs (12.8% vs. 23.3%; PS-adjusted odds ratio [OR] 0.24, 95% confidence interval [CI] 0.06–0.96; Figure 2). The risk of in-hospital death due to rupture was also significantly lower with thoracic INAAs (5.1% vs. 9.3%; PS-adjusted OR 0.11, 95% CI 0.01–0.90). No significant difference was observed in the remaining outcomes. Of note, open conversion was only performed in 1 patient in this cohort and that was a case of a thoracic INAA involving the aortic arch requiring emergency open conversion due to post-TEVAR hemoptysis with suspected aortobronchial fistula formation. Following open total arch replacement, in-hospital death did not occur, and the patient was discharged in a stable condition. No other patients underwent open conversion in this study. However, redo endovascular repair was performed in 5 patients (2 TEVAR, 3 EVAR), with endoleaks being the most common reason for re-intervention. We further compared the in-hospital outcomes of endovascular, open repair and medical therapy (Table 4). Compared with patients who underwent open repair, those with medical therapy alone had a shorter duration of mechanical ventilation and intensive care unit stay (P<0.05).

Table 3. In-Hospital Outcomes

Variable	Stent group			Unadjusted analysis		PS-adjusted analysis
	Total (n=82)	TEVAR (n=39)	EVAR (n=43)	OR (95% CI) of TEVAR	P value	OR (95% CI) of TEVAR	P value
In-hospital death	15 (18.3)	5 (12.8)	10 (23.3)	0.49 (0.15, 1.57)	0.228	0.24 (0.06, 0.96)	0.043
Death due to rupture	6 (7.3)	2 (5.1)	4 (9.3)	0.53 (0.09, 3.05)	0.475	0.11 (0.01, 0.90)	0.039
Reoperation	6 (7.3)	3 (7.7)	3 (7.0)	1.11 (0.21, 5.86)	0.901	1.35 (0.19, 9.67)	0.767
New-onset stroke	3 (3.7)	2 (5.1)	1 (2.3)	2.27 (0.20, 26.07)	0.510	1.19 (0.08, 17.93)	0.902
New onset dialysis	7 (8.5)	3 (7.7)	4 (9.3)	0.81 (0.17, 3.88)	0.795	0.78 (0.13, 4.81)	0.789
Prolonged ventilation	9 (11.0)	4 (10.3)	5 (11.6)	0.87 (0.22, 3.50)	0.843	0.29 (0.06, 1.49)	0.137
Ventilation, days	1 [0, 3]	2 [0, 4]	0 [0, 2]	−0.20 (−4.33, 3.94)	0.925	−2.62 (−7.37, 2.13)	0.276
ICU stay, days	2 [1, 5]	3 [1, 6]	1 [0, 4]	−0.46 (−4.49, 3.57)	0.820	−3.32 (−7.90, 1.26)	0.153
Hospital stay, days	46 [29, 57]	47 [36, 60]	42 [26, 57]	3.74 (−9.27, 16.75)	0.569	1.68 (−13.62, 16.97)	0.828

Data are presented as frequency (percentage) or median [25^th, 75^th percentile]. CI, confidence interval; EVAR, endovascular aortic repair; ICU, intensive care unit; OR, odds ratio; PS, propensity score; TEVAR, thoracic endovascular aortic repair.

Figure 2.

Cumulative incidence (fitted one minus survival curve) of in-hospital death. CI, confidence interval; EVAR, endovascular aortic repair; PS, propensity score; TEVAR, thoracic endovascular aortic repair.

Table 4. In-Hospital Outcomes of Different Treatments

Variables	Stent (n=82)	Open repair (n=29)	Medical (n=22)	P value
In-hospital death	15 (18.3)	4 (13.8)	1 (4.6)	0.323
Death due to rupture	6 (7.3)	2 (6.9)	0 (0.0)	NA
Reoperation	4 (4.9)	0 (0.0)	NA	NA
New-onset stroke	3 (3.7)	1 (3.5)	0 (0.0)	NA
Initiation of dialysis	7 (8.5)	1 (3.5)	0 (0.0)	NA
Prolonged ventilation	9 (11.0)	9 (31.0)	0 (0.0)	NA
Ventilation, days	1 [0, 3]	2 [1, 8]	0 [0, 0]a	0.015
ICU stay, days	2 [1, 5]	5 [2, 15]	0 [0, 0]a	0.002
Hospital stay, days	46 [29, 57]	46 [24, 66]	43 [16, 50]	0.337

Data are presented as frequency (percentage) or median [25^th, 75^th percentile]. ^aSignificant difference vs. the open group. ICU, intensive care unit; NA, not applicable.

Follow-up Outcomes

For those who survived the index surgery, the mean follow-up was 2.5 years (SD 2.8 years). Duration of postoperative antibiotic is provided in Supplementary Table 4. The risk of all-cause death was significantly higher in the thoracic than in the abdominal group (35.3% vs. 15.2%; PS-adjusted hazard ratio [HR] 6.90, 95% CI 1.69–28.19; Figure 3A), with deaths being predominantly infection-related (23.5% vs. 9.1%; PS-adjusted HR 9.71, 95% CI 1.52–62.10; including 6 sepsis, 1 endograft infection and 1 aorto-enteric fistulation; Figure 3B) and due to rupture (11.8% vs. 0%). Please refer to Supplementary Table 5 for details. The 5-year risk of INAA recurrence showed a higher trend with thoracic INAAs (14.7% vs. 3.0%) (Supplementary Table 6). We further analyzed the follow-up outcomes of all treated INAAs (Table 5). Compared with open repair, endovascular therapy and medical therapy alone were associated with more deaths due to rupture (P<0.001). Medical therapy alone was also associated with a higher rate of readmission for aortic disease compared with open repair and endovascular therapy (P<0.001). Blood culture-stratified deaths are shown in Supplementary Table 7.

Figure 3.

Cumulative incidence (fitted one minus survival curve) of all-cause death (A) and infection-related death after discharge (B). CI, confidence interval; EVAR, endovascular aortic repair; PS, propensity score; TEVAR, thoracic endovascular aortic repair.

Table 5. Follow-up Outcomes of Different Treatments for Patients Who Survived to Discharge

Variables	Stent (n=67)	Open repair (n=25)	Medical (n=21)	P value
All-cause death	17 (25.4)	4 (16.0)	7 (33.3)	0.274
Death due to rupture	4 (6.0)	0 (0.0)a	3 (14.3)b	<0.001
Infection-related death	11 (16.4)	3 (12.0)	5 (23.8)	0.509
Reoperation	9 (13.4)	3 (12.0)	9 (42.9)a,b	0.011
Readmission for aortic disease	10 (14.9)	3 (12.0)	14 (66.7)a,b	<0.001

Data are presented as frequency (percentage). ^aSignificant difference vs. the stent group. ^bSignificant difference vs. the open group. ICU, intensive care unit.

Risk Factors for In-Hospital and Total Mortality

In the multivariable logistic regression analysis, the result identified that CKD was a significant risk factor (OR 4.54, 95% CI 1.31–15.75) (Supplementary Table 8). In terms of all-cause death (including in-hospital death and death after surviving to discharge), the multivariable Cox model indicated age (HR 1.04, 95% CI 1.01–1.08) and CKD (HR 3.76, 95% CI 1.80–7.84) as significant prognostic factors (Supplementary Table 9).

Discussion

To our knowledge, this is the first study with comparable sample sizes of endovascularly managed thoracic and abdominal INAAs. The largest cohort to date, published in 2014,⁸ was among the few studies with >100 cases.^8,15,16 However, there were disproportionately more cases of abdominal INAAs in that study. Our cohort can thus provide clinical data more representative of thoracic INAAs. Our study showed that endovascular repair of abdominal INAAs was associated with a higher in-hospital mortality rate than thoracic INAAs, but thoracic INAAs were associated with a high long-term mortality rate, mostly due to infection-related complications. Furthermore, the 5-year risks of INAA recurrence and death were higher with thoracic INAAs. We also demonstrated that Salmonella sp. were the most frequently isolated pathogens in our population, and that fluoroquinolones were the most prescribed agents after 3rd-generation cephalosporins and anti-MRSA agents.

Patient Presentation, Pathogens, and Choice of Antibiotics

As a tertiary referral institution, the proportion of symptomatic patients, as well as the culture-positive rate, was relatively low in our population. Most patients with a negative culture had received partial treatment. Salmonella sp. was the most frequently isolated pathogen in the remaining cases, which is consistent with previous reports from Taiwan.^2,6,17–19 In contrast, gram-positive cocci are more common in Europe,^7–11,16 North America,¹ and in a recent Japanese report.⁴

In accordance with the latest recommendations from the ESVS,¹² IV 3rd-generation cephalosporins and anti-MRSA agents were the most prescribed agents in hospital. Fluoroquinolones were also frequently prescribed. Our team recently discovered that fluoroquinolones are associated with higher risks of all-cause death, aortic death, and later aortic surgery in patients with AA or dissection,^20,21 and further studies are warranted to determine whether the use of fluoroquinolones entails a higher risk for adverse outcomes in INAAs.

Anatomical and Surgical Considerations

Juxtarenal aneurysms and thoracic aneurysms requiring stenting over zones 0–2 are anatomically unsuitable for endovascular repair, frequently mandating adjunctive procedures such as chimney or bypass. Indeed, a juxtarenal location was significantly associated with in-hospital death in our study. The collective results of our cohort suggest that lack of patient selection based on anatomical factors was an important reason for the higher mortality rate. The endovascular approach allows for more aggressive intervention in frail patients who are otherwise poor surgical candidates. A significant number of the study patients were immunocompromised with severe comorbidities and/or infections. Moreover, conversion to open repair was only performed in 1 patient, indicating that despite the traditional belief of open repair being the gold standard for infection control, surgeons at our institution were hesitant to perform it in this fragile patient population. In addition, old age is another important factor that prevents surgeons from adopting the more aggressive open repair approach. However, a recent study reported that performing EVAR solely because of the age of the patient did not improve survival in octogenarians with abdominal AA.²²

Outcomes of TEVAR for Thoracic INAAs

The estimated 30–90-day mortality rate for descending aorta INAAs was 15% for TEVAR and 7–20% for open repair in a previous report.⁷ Scarcity of data has precluded similar analysis of ascending aortic/aortic arch INAAs. Higher short- and long-term mortality rates were observed in our cohort than in a Swedish nationwide cohort of 52 patients, in which the survival rate was 92%, 88%, 78%, and 71% at 30 days, 3 months, 1 year, and 5 years, respectively.¹⁰ Of note, the Swedish cohort consisted predominantly of cases of descending aortic (n=42; 81%) and few cases of aortic arch INAAs (n=6; 11%), whereas more than one-third of the cases involved the aortic arch in our analysis. The apparent poorer outcomes in our study are likely attributable to the unfavorable anatomical locations and aggressive adoption of TEVAR in fragile patients. However, the cause cannot be conclusively determined from this study. The open conversion rate was low in our cohort (only 1 patient underwent open conversion after TEVAR). Patients with a thoracic INAA would likely benefit from elective open repair after initial stabilization with TEVAR. Several studies have demonstrated that elective open conversion after TEVAR for thoracic AA improves morbidity and mortality.^23–26 However, patients with persistent infection have a higher risk of perioperative complications, and careful risk-benefit assessment is crucial.

Outcomes of EVAR for Abdominal INAAs

The estimated 30–90-day mortality rate for infrarenal and iliac INAAs was 3–9% for EVAR and 5–23% for open repair in a previous report.⁷ The in-hospital mortality rate in our study seemed unexpectedly high (e.g., the 3-month survivial was 96% in a Swedish nationwide study⁹). However, in terms of long-term survivial, our results were comparable to those of the Swedish registry study (84% at 1 year), with lower rates of reoperations (24% in the Swedish cohort).⁹

Infection-Related Deaths

In our study, despite the initial favorable outcomes, the long-term mortality rate for TEVAR was high, resulting in a somewhat higher all-cause mortality rate compared with the EVAR group. Infection control was a key determinant of long-term outcomes, which is similar to a previous nationwide study by Sörelius et al,¹⁰ in which short-term survival was good with TEVAR (92%, 88%, 78%, and 71% survival at 30 days, 3 months, 1 year, and 5 years, respectively), but infection-related complications remained a major concern. EVAR for abdominal INAAs, however, was deemed comparable to open repair in several studies.^3,6,8,9 Cautious monitoring is warranted in patients with thoracic INAA managed with TEVAR.

Predictors of Poor Outcomes

In addition to the anatomical location of the aneurysm and age, CKD was identified as an important risk factor for short- and long-term deaths in our cohort. Our team previously demonstrated that in patients undergoing endovascular aortic stent graft therapy, CKD was an independent predictor of higher rates of in-hospital death, postoperative complications, and all-cause death.²⁷ Furthermore, patients on dialysis had the poorest outcomes.²⁷ A recent study of open surgical repair of INAA found similar results, with CKD independently associated with both in-hospital and 90-day deaths.²⁸ Our results further highlighted the significant contribution of renal impairment to the death of patients with INAA. Population aging with a high prevalence of CKD is a significant feature of Taiwan’s and many developed countries. Further studies are required to identify the specific determinants of poor outcome in this patient demographic and the measures to ameliorate such effect.

Study Limitations

Our study was limited by its retrospective, observational nature, which precluded analysis of causality. The population was derived predominantly from tertiary referral centers. Some patients received prior treatment and detailed initial presentation could not be obtained. Lastly, due to geographical variations in pathogens and patients’ characteristics, our results can only reflect the status of INAAs in Taiwan.

Conclusions

In this cohort of patients with INAAs managed with endovascular repair, abdominal INAAs were associated with significantly more in-hospital deaths, and long-term outcomes were worse with thoracic INAAs. CKD was an important predictor of both short- and long-term deaths, and infections were the major cause of death. Preoperative anatomical considerations and early open conversion for infection control might improve outcomes in selected patients.

Acknowledgments

The authors appreciate the support of the Maintenance Project of the Center for Big Data Analytics and Statistics at CGMH (Grant CLRPG3D0049) for statistical consultation. The authors also thank Alfred Hsing-Fen Lin and Ben Yu-Lin Chou of Raising Statistics Consultant Inc. for their statistical assistance.

All authors takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.

Disclosures

The authors report no relationships that could be construed as a conflict of interest.

Funding

This study was supported by Chang Gung Memorial Hospital (CORPG3M0371, CMRPG3L0101, CMRPG3L0102, CFRPG3M0011, BMRPD95 (S.-W.C.)) and the Ministry of Science and Technology (MOST-110-2314-B-182A-114 (S.-W.C.)).

IRB Information

This study was approved by the Institutional Review Board of Linkou Chang Gung Memorial Hospital, Taiwan (IRB No. 201900585B0).

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-23-0146

References

• 1.   Wilson WR, Bower TC, Creager MA, Amin-Hanjani S, O’Gara PT, Lockhart PB, et al. Vascular graft infections, mycotic aneurysms, and endovascular infections: A Scientific Statement from the American Heart Association. Circulation 2016; 134: e412–e460, doi:10.1161/CIR.0000000000000457.
• 2.   Luo CY, Ko WC, Kan CD, Lin PY, Yang YJ. In situ reconstruction of septic aortic pseudoaneurysm due to Salmonella or Streptococcus microbial aortitis: Long-term follow-up. J Vasc Surg 2003; 38: 975–982, doi:10.1016/s0741-5214(03)00549-4.
• 3.   Dang Q, Statius van Eps RG, Wever JJ, Veger HTC. Nationwide study of the treatment of mycotic abdominal aortic aneurysms comparing open and endovascular repair in The Netherlands. J Vasc Surg 2020; 72: 531–540, doi:10.1016/j.jvs.2019.09.060.
• 4.   Hosaka A, Kumamaru H, Takahashi A, Azuma N, Obara H, Miyata T, et al. Nationwide study of surgery for primary infected abdominal aortic and common iliac artery aneurysms. Br J Surg 2021; 108: 286–295, doi:10.1093/bjs/znaa090.
• 5.   Lim YT, Tay WM, Lo ZJ, Pua U, Quek LHH, Tan BP, et al. Endovascular repair of mycotic aortic aneurysms confers good medium-term outcomes and aneurysmal sac resolution. Singapore Med J 2020; 63: 263–267, doi:10.11622/smedj.2020165.
• 6.   Luo CM, Chan CY, Chen YS, Wang SS, Chi NH, Wu IH. Long-term Outcome of endovascular treatment for mycotic aortic aneurysm. Eur J Vasc Endovasc Surg 2017; 54: 464–471, doi:10.1016/j.ejvs.2017.07.004.
• 7.   Sörelius K, Budtz-Lilly J, Mani K, Wanhainen A. Systematic review of the management of mycotic aortic aneurysms. Eur J Vasc Endovasc Surg 2019; 58: 426–435, doi:10.1016/j.ejvs.2019.05.004.
• 8.   Sörelius K, Mani K, Björck M, Sedivy P, Wahlgren CM, Taylor P, et al. Endovascular treatment of mycotic aortic aneurysms: A European multicenter study. Circulation 2014; 130: 2136–2142, doi:10.1161/circulationaha.114.009481.
• 9.   Sörelius K, Wanhainen A, Furebring M, Björck M, Gillgren P, Mani K. Nationwide study of the treatment of mycotic abdominal aortic aneurysms comparing open and endovascular repair. Circulation 2016; 134: 1822–1832, doi:10.1161/circulationaha.116.024021.
• 10.   Sörelius K, Wanhainen A, Wahlgren CM, Langenskiöld M, Roos H, Resch T, et al. Nationwide study on treatment of mycotic thoracic aortic aneurysms. Eur J Vasc Endovasc Surg 2019; 57: 239–246, doi:10.1016/j.ejvs.2018.08.052.
• 11.   Berard X, Battut AS, Puges M, Carrer M, Stenson K, Cazanave C, et al. Fifteen years single-center experience with in situ reconstruction for infected native aortic aneurysms. J Vasc Surg 2021; 75: 950–961.e5, doi:10.1016/j.jvs.2021.08.094.
• 12.   Wanhainen A, Verzini F, Van Herzeele I, Allaire E, Bown M, Cohnert T, et al. European Society for Vascular Surgery (ESVS) 2019 Clinical Practice Guidelines on the management of abdominal aorto-iliac artery aneurysms. Eur J Vasc Endovasc Surg 2019; 57: 8–93, doi:10.1016/j.ejvs.2018.09.020.
• 13.   Shao SC, Chan YY, Yang YHK, Lin SJ, Hung MJ, Chien RN, et al. The Chang Gung Research DatabaseL A multi-institutional electronic medical records database for real-world epidemiological studies in Taiwan. Pharmacoepidemiol Drug Saf 2019; 5: 593–600.
• 14.   Sörelius K, Wanhainen A, Mani K. Infective native aortic aneurysms: Call for consensus on definition, terminology, diagnostic criteria, and reporting standards. Eur J Vasc Endovasc Surg 2020; 59: 333–334, doi:10.1016/j.ejvs.2019.11.008.
• 15.   Jutidamrongphan W, Kritpracha B, Sörelius K, Hongsakul K, Suwannanon R. Features of infective native aortic aneurysms on computed tomography. Insights Imaging 2022; 13: 2, doi:10.1186/s13244-021-01135-x.
• 16.   Sörelius K, Wanhainen A, Furebring M, Mani K, Resch T, Hultgren R, et al. The microbiology of infective native aortic aneurysms in a population-based setting. Ann Vasc Surg 2021; 78: 112–122, doi:10.1016/j.avsg.2021.06.039.
• 17.   Hsu RB, Tsay YG, Wang SS, Chu SH. Surgical treatment for primary infected aneurysm of the descending thoracic aorta, abdominal aorta, and iliac arteries. J Vasc Surg 2002; 36: 746–750, doi:10.1067/mva.2002.126557.
• 18.   Lin CH, Hsu RB. Primary infected aortic aneurysm: Clinical presentation, pathogen, and outcome. Acta Cardiol Sin 2014; 30: 514–521, doi:10.6515/acs20140630a.
• 19.   Yu SY, Hsieh HC, Ko PJ, Huang YK, Chu JJ, Lee CH. Surgical outcome for mycotic aortic and iliac anuerysm. World J Surg 2011; 35: 1671–1678, doi:10.1007/s00268-011-1104-9.
• 20.   Chen SW, Chan YH, Wu VCC, Cheng YT, Chen DY, Lin CP, et al. Effects of fluoroquinolones on outcomes of patients with aortic dissection or aneurysm. J Am Coll Cardiol 2021; 77: 1875–1887, doi:10.1016/j.jacc.2021.02.047.
• 21.   Chen SW, Lin CP, Chan YH, Wu VCC, Cheng YT, Tung YC, et al. Fluoroquinolones and risk of aortic aneurysm or dissection in patients with congenital aortic disease and Marfan syndrome. Circ J 2023, doi:10.1253/circj.CJ-22-0682.
• 22.   Banno H, Sugimoto M, Sato T, Ikeda S, Kawai Y, Tsuruoka T, et al. Endovascular aneurysm repair compared with open repair does not improve survival in octogenarians. Circ J 2021; 85: 2166–2171, doi:10.1253/circj.CJ-21-0574.
• 23.   Coselli JS, Spiliotopoulos K, Preventza O, de la Cruz KI, Amarasekara H, Green SY. Open aortic surgery after thoracic endovascular aortic repair. Gen Thorac Cardiovasc Surg 2016; 64: 441–449, doi:10.1007/s11748-016-0658-8.
• 24.   Joo HC, Kwon JH, Kim JH, Lee SH, Lee S, Youn YN, et al. Late open conversion after thoracic endovascular aortic repair. J Vasc Surg 2019; 70: 439–448.e1, doi:10.1016/j.jvs.2018.11.019.
• 25.   Keschenau PR, Ketting S, Mees B, Barbati ME, Grommes J, Gombert A, et al. Editor’s Choice: Open thoracic and thoraco-abdominal aortic repair after prior endovascular therapy. Eur J Vasc Endovasc Surg 2018; 56: 57–67, doi:10.1016/j.ejvs.2018.03.015.
• 26.   LeMaire SA, Green SY, Kim JH, Sameri A, Parenti JL, Li PH, et al. Thoracic or thoracoabdominal approaches to endovascular device removal and open aortic repair. Ann Thorac Surg 2012; 93: 726–732; discussion 733, doi:10.1016/j.athoracsur.2011.10.080.
• 27.   Wu CC, Chou AH, Lin YS, Wu VC, Chang SH, Chu PH, et al. Late outcomes of endovascular aortic stent graft therapy in patients with chronic kidney disease. Medicine (Baltimore) 2020; 99: e22157, doi:10.1097/md.0000000000022157.
• 28.   Touma J, Couture T, Davaine JM, de Boissieu P, Oubaya N, Michel C, et al. Mycotic/infective native aortic aneurysms: Results after preferential use of open surgery and arterial allografts. Eur J Vasc Endovasc Surg 2022; 63: 475–483, doi:10.1016/j.ejvs.2021.10.041.