Aortoenteric fistula: clinical features, diagnostic challenges, and surgical outcomes — a retrospective analysis of 10 cases

Abstract

Background

Aortoenteric fistula (AEF) is a rare but life-threatening cause of gastrointestinal hemorrhage, often presenting diagnostic challenges and associated with high mortality. Secondary AEF (SAEF), typically related to previous aortic grafting, accounts for most cases. Optimal surgical management and perioperative strategies remain under debate. This study analyzed the clinical characteristics, diagnostic modalities, surgical techniques, and outcomes of patients with AEF to provide insights for guiding future treatment approaches.

Methods

Retrospective analysis of 10 patients surgically treated for AEF at Guangdong Provincial People’s Hospital (2022–2024). Data included demographics, comorbidities, presentation, diagnostics, procedures, complications, and outcomes. All had preoperative imaging and surgery.

Results

Ten patients, all male with a median age of 68 years (interquartile range [IQR] 59–71), comprised 8 SAEF and 2 primary AEF. Fistulas were duodenal (n = 7) and jejunal (n = 3). Presentation: GI hemorrhage (60%), abdominal pain (70%), fever (70%). CTA diagnosed AEF in all 10 patients; EGD (n = 3) was diagnostic in only one. All underwent successful open repair (aneurysm resection, vascular reconstruction, bowel repair). Intraoperative gastroscopy localized the fistula in one case. Postoperative complications occurred in five patients (50%): septic shock (30%), hemorrhage (20%), anastomotic leakage (20%). Two patients (20%) died from severe complications.

Conclusion

In this study, all patients underwent open surgical repair, including aneurysm resection, vascular graft replacement, and intestinal reconstruction. However, postoperative complications were common, and some patients experienced poor outcomes despite timely intervention. These findings highlight the importance of early diagnosis using CTA and prompt surgical management to improve patient prognosis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-025-04539-x.

Background

Aortoenteric fistula (AEF) is an uncommon but life-threatening cause of gastrointestinal (GI) hemorrhage, characterized by abnormal communication between the aorta and the intestinal tract, with its clinical significance lying in its extremely high mortality rate when diagnosis and intervention are delayed [1, 2]. The development of AEF disrupts vascular integrity and may result in massive hemorrhage, making early recognition and urgent surgical management imperative to prevent catastrophic outcomes [3–5]. AEF is classified as primary (PAEF), often resulting from erosion of an atherosclerotic abdominal aortic aneurysm into the bowel [6, 7], or secondary (SAEF), which is more common and typically occurs as a late complication of aortic reconstructive surgery due to graft erosion or infection [8–10]. The incidence rate of PAEF on autopsy is 0.04 to 0.7%, and the post-operative incidence rate of a SAEF is 0.5 to 2.3% [11–13].

The clinical presentation of AEF is often insidious and highly variable. Although the classic triad of GI hemorrhage, abdominal pain and a palpable abdominal mass, has traditionally been recognized as characteristics of AEF [14], these features rarely occur together and are observed in only 6–12% of cases at the time of initial presentation [15, 16]. Instead, patients more commonly exhibit non-specific symptoms such as fever, anemia, or sepsis, which can obscure the diagnosis and delay appropriate management [17, 18]. Moreover, subtle and intermittent hemorrhage is also common and may precede a fatal hemorrhagic event [19], making early identification difficult in routine clinical practice. This non-specific presentation necessitates a high index of suspicion and mandates differentiation from more common causes of GI bleeding and sepsis, such as peptic ulcer disease, angiodysplasia, diverticulosis, malignancies, and inflammatory bowel disease. Particularly in patients with a history of aortic surgery, AEF must be considered a critical—albeit rare—diagnosis in the differential.

Accurate and timely diagnosis of AEF remains challenging. Although esophagogastroduodenoscopy (EGD) is often performed as an initial diagnostic modality, especially in hemodynamically stable patients with GI hemorrhage, its sensitivity remains limited [20]. Factors such as the anatomical location of the fistula, the intermittent nature of hemorrhage, and the presence of submucosal tracts may prevent direct visualization during endoscopy [8, 21–23]. Due to these limitations, computed tomography angiography (CTA) is now considered the preferred imaging modality for diagnosing AEF as it provides high sensitivity and specificity and enables the detection of key radiological features such as ectopic gas, peri-graft fluid collections, and contrast extravasation [24, 25]. In addition, CTA offers valuable information for preoperative planning by delineating the extent of vascular and intestinal involvement [12, 26, 27].

Surgical intervention remains the primary treatment for AEF and is the only definitive treatment modality, as without surgical repair, mortality may approach 100%^4,28. Open surgery is generally preferred in patients who are suitable for surgery, as it allows complete removal of infected graft components and resection of the involved bowel segment. In patients who are critically ill or considered high-risk for open surgery, endovascular approaches may be considered as an alternative. Although endovascular aortic repair (EVAR) may improve short-term survival, it is associated with a risk of persistent infection and its long-term outcomes remain uncertain [28–32].

Herein, the present study systematically evaluated the clinical characteristics, surgical strategies, and outcomes of patients with AEF treated at our institution. We hypothesized that a comprehensive surgical approach encompassing radical debridement, vascular reconstruction, bowel resection with anastomosis, and adjunctive techniques such as omental wrapping would reduce postoperative infectious complications and improve survival. Additionally, we aimed to clarify the clinical presentation, diagnostic pathways, and intraoperative decision-making involved in managing this complex condition to provide practical insights that may inform future clinical practice.

Methods

Study design and patient selection

A retrospective analysis was conducted on patients with AEF who underwent open surgical treatment at Guangdong Provincial People’s Hospital between January 2022 and December 2024. The study protocol for collecting clinical specimens was approved by the Institutional Ethics Committee of Guangdong Provincial People’s Hospital (Guangdong Academy of Medical Sciences) under the approval number KY2024-892-01. The study inclusion criteria were: (1) diagnosis of AEF confirmed intraoperatively or through preoperative imaging; (2) completion of definitive open surgical management at our institution. Patients without surgical intervention or with incomplete medical records were excluded. The diagnostic and therapeutic algorithm for AEF are summarized in Fig. 1.

Fig. 1.

Diagnostic and therapeutic algorithm for AEF

Preoperative assessment and diagnosis

AEF was defined as a pathological connection between the aorta and the GI tract. Due to its frequently non-specific and insidious clinical presentation, early diagnosis is often difficult. Therefore, all patients underwent contrast-enhanced CTA as the primary diagnostic modality. CTA was performed using multidetector scanners with thin-slice acquisition (≤ 1 mm) after intravenous contrast administration. The diagnosis of AEF on CTA was considered positive when at least one of the following direct or indirect signs was present: (1) ectopic gas (pneumatosis) in the periaortic or perigraft region; (2) perigraft fluid collections or soft tissue thickening; (3) extravasation of intravenous contrast into the bowel lumen; (4) loss of the normal fat plane between the aorta and the adjacent bowel wall; (5) pseudoaneurysm formation at the suture line or graft site. All CTA studies were independently interpreted by two experienced cardiovascular radiologists (each with > 10 years of experience in vascular imaging), who were blinded to the endoscopic findings. In selected patients presenting with upper GI hemorrhage but inconclusive CTA findings, EGD was performed to evaluate for mucosal defects or active hemorrhage. EGD was not routinely performed in patients with suspected AEF. Its use was reserved for hemodynamically stable patients with upper GI hemorrhage of unclear origin after initial CTA, or when CTA findings were equivocal. EGD was considered diagnostic for AEF only if a pulsatile mass, adherent clot in the duodenum, or a direct fistulous opening was visualized. EGD was deemed non-diagnostic if it revealed only nonspecific findings or failed to identify the source of bleeding. Additional routine assessments included abdominal ultrasonography, plain abdominal radiography, and standard laboratory tests. Intraoperative gastroscopy was selectively employed in cases where the anatomical relationship between the fistula and surrounding structures was unclear, to assist in localizing the fistulous opening and guiding surgical repair.

Surgical procedures

The surgical management of AEF at our institution was guided by a standardized protocol implemented by a multidisciplinary team (MDT) comprising vascular surgeons, gastrointestinal surgeons, infectious disease specialists, and radiologists. Upon diagnosis, all patients underwent immediate resuscitation. Our institutional strategy prioritized definitive open surgical repair (OSR) with radical debridement as the treatment of choice for all patients deemed surgically fit. EVAR was not employed as a definitive therapy in this cohort. Its role was strictly limited to that of a “bridge to surgery” in a specific, high-risk scenario: for patients presenting with ongoing, life-threatening hemorrhagic shock who were too unstable to tolerate immediate OSR. In such cases, EVAR served to achieve immediate hemorrhage control, allowing for subsequent stabilization and a planned, single-stage OSR within 48–72 h.

The open surgical procedure commenced with radical debridement of all infected periaortic tissues, resection of the fistulous tract, and complete removal of any pre-existing infected prosthetic graft. In-situ reconstruction (ISR) was our standard and preferred approach for vascular continuity. Concurrently, the management of the intestinal tract involved mandatory resection of the involved bowel segment. Simple suture repair was avoided. Intestinal continuity was restored via either a duodenojejunostomy for fistulas involving the third or fourth portion of the duodenum, or a duodenoileal anastomosis for those at the duodenal-jejunal flexure or more distal sites. The specific anastomotic configuration (side-to-side or end-to-side) was determined intraoperatively based on surgical judgment to ensure a tension-free and well-vascularized anastomosis. Finally, an omental pedicle flap was routinely mobilized and interposed between the newly placed vascular graft and the intestinal suture line to serve as a vital biological barrier and to enhance local healing.

Microbiological analysis

Specimens, including peripheral blood, purulent exudates from infected sites and intraoperative tissue samples, were collected under sterile conditions for microbiological evaluation. Blood cultures were obtained preoperatively and intraoperative samples were collected directly from areas of graft infection, abscess cavities, and adjacent bowel tissue. To distinguish true infection from contamination or colonization, a positive culture was considered clinically significant only when it was confirmed by growth from multiple independent sample types or by heavy and repeated growth from a sterile site. All specimens were processed immediately using standard microbiological protocols [5]. The cultures were incubated under aerobic and anaerobic conditions, and pathogen identification was performed. Antimicrobial susceptibility testing was conducted and fungal cultures were also performed where clinically indicated [33, 34].

Postoperative management

Postoperative care was standardized and initiated immediately upon transfer to the intensive care unit (ICU). Hemodynamic monitoring included continuous ECG, invasive arterial pressure, central venous pressure, and urine output measurements. Empiric broad-spectrum antibiotics (e.g., piperacillin-tazobactam or meropenem with vancomycin) were initiated during surgery and subsequently adjusted based on intraoperative culture results and susceptibility profiles.

Nutritional support was individualized; patients unable to tolerate oral intake received early enteral nutrition via nasogastric or nasojejunal tubes. The timing and type of nutritional support (elemental, semi-elemental, or polymeric) were determined based on GI recovery and tolerance. The duration of nasogastric tube (NGT) retention, ICU stay, and overall hospital stay were documented in all cases.

In cases of postoperative complications such as sepsis, anastomotic leakage, or hemorrhage, targeted interventions, including antimicrobial escalation or drainage procedures were performed as clinically indicated.

Antimicrobial Therapy

Upon clinical suspicion of AEF-associated sepsis, broad-spectrum empirical antibiotics were initiated intraoperatively. The standard regimen included coverage for Gram-negative bacteria and Gram-positive bacteria. The empirical regimen was subsequently de-escalated or adjusted based on intraoperative culture results and antimicrobial susceptibility profiles. The final targeted regimen was tailored to the isolated pathogens and continued for a total duration of 4 to 6 weeks postoperatively, depending on the clinical response and control of the infectious focus. For cases with confirmed fungal infection systemic antifungal therapy was initiated. The choice of agent was guided by the isolated species and susceptibility data. Initial intravenous therapy was administered for at least 2 weeks or until clinical stability was achieved. This was followed by a step-down to oral antifungal agents to complete a total course of at least 6–8 weeks of antifungal therapy. All antimicrobial regimens were tailored with the assistance of experienced infectious disease physicians.

Outcome assessment and follow-up

Operative records were reviewed to collect intraoperative data, including blood loss (quantified via suction canisters and gauze count) and total operative time (from incision to closure). Postoperative complications were recorded and major complications such as cerebral infarction, pulmonary edema, hemorrhage, intestinal leakage, and anastomotic leakage were diagnosed via imaging, clinical presentation, and surgical findings.

Patient outcomes were assessed through both inpatient records and follow-up. Follow-up was conducted via scheduled outpatient visits or telephone interviews and focused on survival status, fistula recurrence, readmissions, and delayed complications. The duration of follow-up was calculated from the date of surgery to the last documented clinical contact or death.

Statistical analysis

Descriptive statistical analysis was performed using SPSS software (version 25.0; IBM Corp., Armonk, NY) Continuous variables were assessed for distribution and expressed as medians with interquartile ranges (IQR), depending on data normality. Categorical variables were reported as counts and percentages. No inferential statistical tests were applied due to the small sample size and descriptive nature of the study.

Results

Background and clinical parameters

A total of 10 patients with AEF were included in this study. The background and clinical parameters of all 10 patients are summarized in Table 1. The cohort consisted entirely of male patients, with a median age of 68 years (IQR 59–71) and a median body mass index of 24 kg/m² (IQR 20–27). For these eight patients with SAEF, the median time interval between the initial aortic graft placement and the diagnosis of AEF was 31.5 months (IQR, 14-69.5 months). Two patients (Cases 1 and 9) presented with acute hemorrhagic anemia (AHA), among whom one (Case 9) progressed to hemorrhagic shock (HS). Comorbidities were highly prevalent, with hypertension present in 90% of patients, coronary heart disease in 30%, and diabetes mellitus in 20%. Abscess formation (AF) was observed in two patients (Cases 2 and 4). Less common comorbidities included a history of lung cancer and chronic obstructive pulmonary disease, each occurring in one patient.

Table 1.

Background and clinical parameters of 10 patients with AEF

Case	Age/Sex	BMI	Comorbidities	Anamnesis	PAEF/SAEF	Graft-to-AEF interval (months)	AEF location (bowel)	Cause	Preoperative workup
1	59/M	20	HS, AHA	CHD, DM, HTN	SAEF	27	Jejunum	Graft infection	CTA, EGD, X-ray, ECG
2	71/M	24	AF	DM, HTN	SAEF	36	Duodenum	Graft infection	CTA, X-ray, ECG, CDE
3	61/M	27	-	HTN	PAEF	-	Jejunum	AAA	CTA, X-ray, ECG
4	71/M	15	AF	HTN	SAEF	4	Duodenum	Graft infection	CTA, X-ray, ECG, CDE
5	71/M	27	-	HTN	SAEF	11	Jejunum	Graft infection	CTA, EGD, X-ray, ECG, CDE
6	66/M	17	-	CHD, HTN, COPD	SAEF	55	Duodenum	Graft infection	CTA, X-ray, PET/CT, ECG
7	70/M	24	-	CHD, HTN, PLC	SAEF	192	Duodenum	Graft infection	CTA, X-ray, PET/CT, ECG, CDE
8	72/M	21	-	HTN	SAEF	84	Duodenum	Graft infection	CTA, X-ray, ECG, CDE
9	50/M	27	AHA	-	PAEF	-	Duodenum	AAPA, infection	CTA, X-ray, ECG
10	58/M	27	-	HTN	SAEF	17	Duodenum	Graft infection	CTA, EGD, X-ray, ECG

Aortoenteric fistula (AEF)

Sex: male (M)

Body Mass Index (BMI)

Comorbidities: Hemorrhagic Shock (HS); Acute Hemorrhagic Anemia (AHA); Abscess Formation (AF)

Anamnesis: Coronary Heart Disease (CHD); Diabetes Mellitus (DM); Hypertension (HTN); Chronic Obstructive Pulmonary Disease (COPD); Postoperative lung cancer (PLC)

PAEF/SAEF: Primary aortoenteric fistula (PAEF); Secondary aortoenteric fistula (SAEF)

Cause: Abdominal Aortic Aneurysm (AAA)

Preoperative workup: Computed Tomography Angiography (CTA); Electrocardiogram (ECG); Esophagogastroduodenoscopy (EGD); Color Doppler Echocardiography (CDE); Positron Emission Tomography/Computed Tomography (PET/CT)

Etiologically, secondary AEF (SAEF) accounted for 80% of cases, all associated with prior aortic graft infection, while primary AEF (PAEF) was identified in the remaining two patients. The duodenum was the most frequently involved intestinal segment (70%), followed by the jejunum (30%).

The underlying causes of AEF were diverse. Graft infection was the most common cause and was identified in eight patients (Cases 1, 2, 4–8, and 10). One case (Case 3) was attributed to an AAA, while another (Case 9) involved a combination of atherosclerotic abdominal aortic aneurysm (AAPA) and infection.

Preoperative assessments were comprehensive and included multiple diagnostic modalities. CTA was performed in all patients (10/10, 100%) and served as the primary diagnostic tool. Routine evaluations included chest radiography and electrocardiography (ECG) in all patients. EGD and color Doppler echocardiography (CDE) were performed in three patients each (30%). Hybrid positron emission tomography/computed tomography (PET/CT) was used in two patients (20%) to assist in identifying suspected graft infections [35].

Clinical features and laboratory testing

The clinical characteristics and laboratory findings of all 10 patients are summarized in eTable 1. The most frequent clinical presentations included abdominal pain (70%), fever (70%), and gastrointestinal hemorrhage (60%). Systemic infection or sepsis was present in 90% of patients, while constitutional symptoms such as weight loss and anorexia were noted in 30% of cases. Symptom duration varied widely, from acute onset within 24 h to a chronic course lasting up to three years.

Laboratory testing commonly revealed anemia (90%) and leukocytosis (50%). Inflammatory markers were markedly elevated, with elevated C-reactive protein observed in 80% of tested patients and elevated procalcitonin in 87.5%.

Infection or sepsis was present in all patients except Case 3, who instead presented with a palpable pulsatile abdominal mass. Escherichia coli was the most commonly isolated pathogen (30%). Other identified organisms included Streptococcus constellatus, Enterococcus faecium, Klebsiella pneumoniae, Candida albicans, Streptococcus sanguinis, and Brucella species.

Imaging results

Computed tomography angiography (CTA) successfully identified all cases of AEF, revealing several characteristic imaging features. The most frequent findings included periartificial effusion and gas accumulation (60%), observed in six patients (Fig. 2), and inflammatory changes such as fatty interface blurring (40%) and tissue infiltration (40%). Direct evidence of active fistulization, indicated by intravascular contrast extravasation into the bowel lumen or surrounding tissues, was present in four patients (Fig. 3). A poorly defined interface between the aortic aneurysm and adjacent intestine, suggestive of erosion or fistula formation, was noted in two cases.

Fig. 2.

Abdominal computed tomography (CT) of case 8. Arrow: CT showed the false lumen outside the stent-covered area is thrombus-filled, with multiple scattered gas shadows inside. Red line: The boundary between the anterior wall of the abdominal aortic aneurysm and the horizontal segment of the duodenum is indistinct

Fig. 3.

Abdominal computed tomography (CT) of case 9. Arrow: CT showed abdominal aortic pseudoaneurysm with perianeurysmal hematoma formation, complicated by infection and rupture into the horizontal segment of the duodenum

Additional CTA findings comprised thrombus within vascular grafts (40%), aortic wall calcification (universal), aortic ulceration (50%), and type II endoleak in two patients with prior endovascular repair. In contrast, esophagogastroduodenoscopy (EGD), performed in three patients, was diagnostic in only one case, directly visualizing a duodenal fistula (Fig. 4), and the remaining two EGD examinations revealed only chronic superficial gastritis. A complete summary of imaging findings is provided in Table 2.

Fig. 4.

Duodenoscopy findings in case 10. Arrow: Duodenoscopy revealed an exposed stent 1 cm distal to the duodenal papilla, covered with yellowish-white purulent exudate, suggesting an aorto-duodenal fistula

Table 2.

Imaging results of patients with AEF

Preoperative workup	Imaging results	Cases
CTA	Periartificial blood vessel effusion and gas accumulation	1, 2, 4, 5, 8, 10
	​Fatty interface blurring	4, 6, 9, 10
	Infection with abscess formation	2, 4, 6
	Inflammatory infiltration involving surrounding tissues	2, 4, 5, 6
	Intravenous contrast within the GI lumen or around the aorta	2, 5, 7, 9
	Fuzzy demarcation between the periaortic aneurysm and the intestinal tract	8, 9
	Thrombus formation in artificial blood vessels	1, 5, 7, 8
	Calcification of the arterial wall	All cases
	Aortic ulcer/ Penetrating aortic ulcer	2, 3, 4, 5 ,7
	​Type II endoleak	5, 7
EGD	Chronic superficial gastritis	1, 5
	Duodenal fistula	10

 Computed Tomography Angiography (CTA); Esophagogastroduodenoscopy (EGD); Gastrointestinal (GI)

Treatments and outcomes

All patients underwent open surgical repair comprising aneurysm resection, in-situ vascular graft replacement, and bowel resection with anastomosis (eTable 2). The median aneurysm diameter was 1.5 cm (IQR 1–2), and intraoperative purulence was observed in 7 patients (70%). Reconstruction was achieved via duodenojejunostomy in 8 patients (80%) and duodenoileal anastomosis in 2 (20%), with side-to-side configuration employed in 6 cases (60%).

The procedures were substantial, with a median intraoperative blood loss of 1,250 mL (IQR 800-2,000) and a median operative time of 418.5 min (IQR 410–446). All patients required blood transfusion and postoperative ICU care. Recovery metrics showed a median nasogastric tube retention of 9.5 days (IQR 7–21), median ICU stay of 4.5 days (IQR 2-6.75), and median total postoperative hospitalization of 18 days (IQR 16–25).

Postoperative complications of varying severity occurred in five patients (50%). For instance, Case 1 developed neurological symptoms on postoperative day 10 (POD10), and cranial CT confirmed cerebral infarction. The patient received targeted symptomatic treatment and was discharged one week later with full recovery.

Case 5

had a history of long-standing infection and prior conservative management. Although there was an initial postoperative improvement in infectious markers, the patient subsequently developed persistent pyrexia, indicating relapse. An adjusted antimicrobial regimen combined with nasogastric decompression led to gradual clinical improvement and eventual uneventful discharge.

Case 6

experienced a severe postoperative course. On POD1, the patient underwent emergency reoperation for intraperitoneal hemorrhage and achieved temporary hemodynamic stabilization. However, on POD10, the patient developed hypotension and tachypnea consistent with septic shock, accompanied by copious brownish peritoneal drainage and formation of an intestinal leakage. A third surgery was performed on POD15. Despite aggressive resuscitative and surgical efforts, the patient progressed to multiorgan dysfunction syndrome (MODS), and life-sustaining therapy was withdrawn on POD28 due to poor prognosis.

Case 7

developed pulmonary edema and cerebral infarction on POD7. With intensive medical treatment, both complications resolved completely, and the patient was successfully discharged on POD 12. Case 8 initially presented with colicky abdominal pain and vomiting on POD16, which was managed conservatively with nasojejunal feeding. However, the patient was readmitted on POD39 due to acute abdominal pain, nausea, and vomiting. Abdominal CT revealed an intestinal leakage and partial intestinal obstruction. Emergency surgery was performed on POD46, but the patient died intraoperatively due to massive hemorrhage of gastrointestinal tract and refractory disseminated intravascular coagulation (DIC).

Discussion

This retrospective study revealed several important findings. First, SAEF was the predominant etiology, primarily associated with previous aortic grafting. The duodenum was the most frequently involved intestinal segment. Second, GI hemorrhage, fever, and abdominal pain were common but non-specific symptoms, often contributing to delayed diagnosis. Third, CTA demonstrated greater diagnostic utility than EGD, which frequently failed to identify the fistulous tract. Lastly, although open surgical repair was effective in controlling the disease, postoperative morbidity and mortality remained significant, mainly due to infectious and GI complications.

The clinical presentation of AEF is often subtle and heterogeneous, making timely diagnosis difficult. In this series, only one patient presented with the classical triad of GI hemorrhage, abdominal pain, and a palpable abdominal mass. Most patients exhibited non-specific findings such as sepsis or anemia, consistent with previous reports describing the diagnostic challenge of AEF [4, 36, 37]. Moreover, clinicians may misattribute these symptoms to more common causes, particularly in patients with a history of aortic grafting, further emphasizing the importance of clinical vigilance [38, 39], thereby emphasizing the importance of maintaining a high index of suspicion, particularly in patients with a history of aortic surgery and unexplained infection or hemorrhage.

Abdominal pain in patients with AEF is believed to arise from localized pathological changes at the site of the fistula. These include mechanical irritation caused by erosion of the aortic wall or pressure exerted by the graft on adjacent bowel tissue, which may activate visceral pain pathways [36, 40]. In addition, leakage of intestinal contents into the peritoneal or retroperitoneal space can induce chemical peritonitis, presenting clinically as persistent or colicky abdominal pain [37]. Since these symptoms often mimic those of more common abdominal conditions, especially in individuals with a history of aortic grafting, they are frequently misdiagnosed and may delay the identification of the underlying fistula. These findings highlight the importance of maintaining a high index of suspicion in patients with unexplained sepsis or GI hemorrhage who have undergone prior aortic surgery [38, 39].

Our findings, consistent with the literature, strongly suggest that when AEF is clinically suspected, urgent CTA should be the first-line investigative tool and should supersede routine EGD. This approach can prevent critical delays in diagnosis and treatment, as EGD often fails to visualize the fistula and may provide a false sense of security,, particularly in cases where fistulas are located in anatomically challenging regions or present with intermittent hemorrhage [20, 41]. Several factors contribute to this low diagnostic yield. First, the fistulous opening may be situated in locations that are difficult to visualize during standard endoscopy, such as the distal duodenum or the graft-enteric interface. Second, hemorrhage from AEF is often intermittent or concealed by clot formation, further limiting direct visualization. Additionally, submucosal tunneling between the aorta and intestinal lumen may obscure the fistulous tract, making it difficult to detect endoscopically [8, 21–23]. In our cohort, preoperative EGD identified only one case of duodenal fistula, while two cases were misinterpreted as chronic gastritis. Despite these limitations, intraoperative endoscopy proved helpful in selected situations. In Case 10, intraoperative EGD successfully localized the fistula orifice, aided in preserving the duodenal papilla, and facilitated surgical planning. In contrast, CTA demonstrated high diagnostic reliability in preoperative assessment. CTA effectively identified key features such as peri-graft gas, fluid collections, and indistinct borders between the aorta and adjacent bowel, all of which contributed significantly to diagnostic accuracy and surgical decision-making [12, 24, 42, 43]. Although highly specific signs like contrast extravasation were observed in only a subset of patients, CTA remained essential in guiding operative strategy.

Surgical repair remains the cornerstone of AEF management. In our study, all patients underwent OSR, which allowed for complete debridement of infected tissue and definitive intestinal reconstruction. This approach was favored at our institution whenever the patient’s condition allowed. Prior studies have shown that OSR offers superior infection control compared to endovascular techniques, particularly in eliminating infected graft material and reducing the risk of late-onset sepsis [27–29]. While EVAR has been associated with better short-term survival in selected SAEF patients, concerns remain regarding its long-term durability and the risk of persistent or recurrent infection [28, 30, 37]. Accordingly, in our institution, OSR was prioritized whenever the patient’s condition permitted it. Procedures performed by general surgeons with greater familiarity in GI surgery may improve technical precision in bowel resection and reconstruction, potentially leading to better outcomes [17].

GI tract management plays a critical role in AEF treatment. In our series, simple repair techniques such as duodenorrhaphy were avoided in favor of bowel resection and anastomosis, consistent with evidence supporting better outcomes with this approach [17, 36]. Omental interposition was routinely applied to separate the graft from the bowel, reduce infection risk, and improve healing. This technique, previously associated with lower mortality and recurrence rates, was particularly beneficial in cases with localized infection and minimal purulence [17, 44]. Additionally, limiting graft excision to infected segments likely minimized operative trauma and supported recovery.

Despite comprehensive surgical strategies, postoperative complications remained frequent and often fatal. Septic shock, hemorrhagic events, and intestinal leakage were the primary contributors to mortality, consistent with previous reports identifying these as the most difficult challenges in AEF management [9, 27]. These outcomes reflect the complexity of treating patients with severe intraoperative contamination and multiple comorbidities, underscoring the importance of multidisciplinary perioperative care. In our cohort, supportive measures such as nasogastric decompression, targeted antimicrobial therapy, and intensive care unit support were critical in reducing the risk of deterioration [45]. Nevertheless, the prognosis of AEF remains poor. Although our protocols may have improved infection control and reduced reinfection, the high incidence of severe complications highlights the need for further optimization. GI complications, in particular, continue to jeopardize postoperative survival [3, 27]. Future efforts should focus on enhancing perioperative infection control, refining GI management strategies, and improving patient selection and risk stratification. While open repair remains the standard treatment, the role of EVAR is evolving. In selected patients with limited infection or high operative risk, EVAR may offer short-term survival benefits. However, its long-term effectiveness remains uncertain due to concerns about persistent infection, warranting further investigation [46].

Our institutional experience, as detailed in this series, reinforces several established principles of AEF management. Beyond this, it offers a refined and systematic approach that may contribute to improved outcomes. The strategic use of intraoperative endoscopy to safeguard critical structures like the duodenal papilla during duodenal resection represents a technical nuance that, to our knowledge, has not been sufficiently emphasized in the literature. Furthermore, our consistent outcomes, achieved through a standardized protocol of radical debridement, in-situ reconstruction with antibiotic-soaked graft, mandatory bowel resection, and routine omental interposition, provide a validated model for a comprehensive “surgical package”. This integrated approach, managed by a multidisciplinary team, ensures complete source control and anatomical reconstruction, which we believe is a key contributor to the successful management of this complex condition.

Several limitations of this study should be acknowledged. First, its retrospective nature and single-center design, along with the small sample size typical of this rare condition, may limit the generalizability of the findings. Furthermore, as our cohort included only patients who underwent open surgical repair, it is subject to a significant selection bias toward those who were either diagnosed early or were hemodynamically stable enough to be considered surgical candidates. This likely explains the relatively high proportion of stable patients on admission and means our outcomes may not reflect those of the most critically ill AEF patients who die before diagnosis or surgery. Second, heterogeneity in patient characteristics and surgical strategies may have introduced selection and treatment bias. Third, the lack of long-term follow-up data prevented evaluation of late complications such as graft reinfection or recurrent fistula formation.

Conclusion

In conclusion, AEF is a rare but life-threatening condition that requires timely diagnosis and urgent surgical intervention. However, diagnosis is often delay due to non-specific symptoms and inconspicuous imaging findings. To improve outcomes, clinicians must maintain a high level of suspicion in at -risk patients and promptly pursue appropriate imaging and surgical management when AEF is suspected. In addition, we highlight the value of specific technical adjuncts, such as intraoperative endoscopy for complex fistula localization, and the consistent application of a comprehensive approach encompassing vascular and enteric reconstruction with biological reinforcement. These refinements in surgical technique and perioperative protocol are essential to improving survival in this high-risk population.

Abbreviations

AAA

Abdominal aortic aneurysm

AAPA

Atherosclerotic abdominal aortic aneurysm

AEF

Aortoenteric fistula

AF

Abscess formation

AHA

Acute hemorrhagic anemia

BP

Blood pressure

BMI

Body mass index

CDE

Color Doppler echocardiography

CHD

Coronary heart disease

COPD

Chronic obstructive pulmonary disease

CRP

C-reactive protein

CTA

Computed tomography angiography

DA

Duodenoileal anastomosis

DJ

Duodenojejunostomy

DM

Diabetes mellitus

ECG

Electrocardiography

EGD

Esophagogastroduodenoscopy

EVAR

Endovascular aortic repair

GI

Gastrointestinal

HGB

Hemoglobin

HS

Hemorrhagic shock

HTN

Hypertension

ICU

Intensive care unit

MODS

Multiorgan dysfunction syndrome

NGT

Nasogastric tube

OSR

open surgical repair

PAEF

Primary aortoenteric fistula

PCT

Procalcitonin

PET/CT

Positron emission tomography/computed tomography

PLC

Postoperative lung cancer

POD

Postoperative day

SAEF

Secondary aortoenteric fistula

SD

Standard deviation

WBC

White blood cell count

Authors’ contributions

J.W. participated in data analysis and interpretation and article drafting. Y.Z. participated in data analysis and interpretation and article drafting. S.C. participated in data analysis and interpretation and article drafting. S.S. participated in data analysis and interpretation and in critically revising the manuscript for important intellectual content. C.K. participated in the study design, data analysis and interpretation.W.H. participated in data analysis and interpretation and article drafting. C.L. participated in data analysis and interpretation and article drafting. Z.L. participated in data analysis and interpretation and article drafting. J.S. participated in data analysis and interpretation and in critically revising the manuscript for important intellectual content. C.H., X.Y. and Z.L. participated in the study design, data analysis and interpretation and gave final approval for publication All authors read and approved the final manuscript.

Funding

This work was supported by Leading Innovation Specialist Support Program of Guangdong Province, the Science and Technology Planning Project of Ganzhou (No. 202101074816), and National Natural Science Foundation of China (No. 82403570).

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due to their containing information such as indirect identifiers that may compromise the privacy of research participants, but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol for collecting clinical specimens was approved by the Institutional Ethics Committee of Guangdong Provincial People’s Hospital (Guangdong Academy of Medical Sciences) under the grant number of KY2024-892-01. The procedures used in this study adhere to the tenets of the Declaration of Helsinki. Informed consent was obtained from all individual participants included in the study.

Consent for publication

All participants provided written informed consent for publication.

Competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.