Background

A secondary aortoenteric fistula (SAEF) is an abnormal connection between the aorta and gastrointestinal tract resulting from reconstructive surgery for an abdominal aortic aneurysm (AAA), including open repair surgery and endovascular treatment. Although rare (incidence rate, 1.6 to 4%) [1,2,3,4], it is life-threatening and has a high mortality rate (24 to 45.8%) [5,6,7]. Elucidating the clinical features and improving the diagnosis of SAEF would therefore be valuable.

The mechanism underlying the pathological development of SAEFs remains unknown. Two hypotheses have been postulated. One suggests an origin in continuous stimulation due to aortic pulsation directly affecting the walls of the intestinal tract and arteries [2, 4, 7,8,9]. This theory is supported by the fact that most SAEFs involve the third or fourth portion of the duodenum which are compressed between the superior mesenteric artery and abdominal aorta in the retroperitoneal space [5]. The other argues an origin in a local inflammatory response due to prosthesis infection during primary surgery [5, 7,8,9,10]. Several species of bacteria not usually identified in the intestine, such as Staphylococcus, have been detected in aortic prostheses in cases of SAEF, which supports this hypothesis.

Here, we describe an aortoenteric fistula following reconstructive surgery for an AAA and discuss the implications of postmortem pathological analysis. We believe the findings further our understanding of the underlying mechanism of SAEFs, which could be useful in deciding a preventive strategy.

Written informed consent was obtained from the patient’s family for publication of this case.

Case presentation

A 59-year-old Japanese man presented to our hospital with the chief complaint of hematochezia and malaise. On the day of admission and 10 days earlier, he had produced a fresh bloody stool. He had undergone open surgery with a bifurcated graft for an AAA 20 months earlier. The course was uneventful, with no remarkable findings on computed tomography (CT) at 6 and 18 months postoperatively. An abdominal examination at our hospital revealed nothing remarkable and no tenderness. His blood pressure was 122/75 mmHg; heart rate, 86/minute; body temperature, 36.6 °C; breathing, 16 per minute; and hemoglobin level, 9.0 g/dL.

Esophagogastroduodenoscopy and total colonoscopy revealed only colon diverticula and no bleeding. Contrast-enhanced CT revealed gas within the aneurysm sac (Fig. 1a, b, yellow arrowhead) and adhesion between the graft and intestinal tract in three areas: the ileum had attached to the anastomosis between the left branch of the graft and left common iliac artery (Fig. 1a, yellow circle); the jejunum to the middle of the graft body; and the duodenum to the anastomosis between the aorta and the proximal graft. Enterococcus faecium was isolated from blood culture, suggesting communication between the intestinal tract and aorta at the attached sites, possibly due to infection of the graft. His vital signs were stable, so surgery was scheduled to take place after antibiotic treatment. After admission, he produced another fresh bloody stool, but bleeding ceased immediately. At 18 days after the second fresh bloody stool, however, he suddenly went into a state of shock, with massive fresh bloody stool and hematemesis, followed by cardiac arrest. Despite intensive cardiopulmonary resuscitation, he died from hemorrhagic shock.

Fig. 1
figure 1

Contrast-enhanced computer tomography (coronary (a) and axial (b) view) revealed gas within aneurysm sac (yellow arrowhead), and attachment of ileum to anastomosis between left branch of graft and left common iliac artery (yellow circle)

Full size image

An autopsy performed with written consent from the family revealed an ulcerative fistula in the distal ileum that adhered to the anastomosis between the left branch of the graft and the left common iliac artery (Fig. 2a), with a small hole at the aortic anastomosis (Fig. 2b). Arterial structure was destroyed at the anastomotic site. There was fibrous thickening of the arterial wall (Fig. 3a, b) and the external elastic lamina had disappeared. The serosa of the small intestine and adventitia of the artery were firmly adhered (Fig. 3c, d). There was no marked sign of infection, such as inflammatory cells or phagocytosis, even around the fistula.

Fig. 2
figure 2

Image at autopsy. Adhesion of the distal ileum to the anastomosis between the left branch of the graft and the left common iliac artery (arrow), and the postoperative aneurysm sac on the head side (asterisk) (a), with a small hole at the aortic anastomosis in dissected ileum (b)

Full size image
Fig. 3
figure 3

Histology of left common iliac artery by Elastica van Gieson stain revealed fibrotic and thickened intima of anastomosed artery exposed to intestinal lumen at edge of fistula (a whole, magnification × 1; b square, magnification × 400). Histology of fistula revealed firm adhesion (arrow) between serosa of intestinal tract and adventitia of artery (asterisk) (c whole, magnification × 1; d square, magnification × 40)

Full size image

Discussion

SAEFs are a rare but life-threatening complication after aortic aneurysm reconstruction [1,2,3, 11]. Despite the importance of establishing the underlying mechanism of fistula formation, their etiology remains unclear. It has been suggested that either chronic infection of the graft or physical stimulation, such as from aortic pulsation pressure, contributes to their formation, which is usually determined based solely on the clinical course or surgical findings [12,13,14,15,16,17,18,19]. To the best of our knowledge, only 25 case reports on SAEFs have been published (see Table 1). Of these, only six discuss the possible mechanism of fistula formation and the results of pathological examination [20,21,22,23,24,25]; and only two of these reports include detailed pathological images [20, 25]. Therefore, the precise mechanism underlying fistula formation remains to be clarified pathologically.

Table 1 Summary of published case reports describing secondary aortoenteric fistula after both open surgery and endovascular aortic reconstruction for abdominal aortic aneurysm
Full size table

The present autopsy findings suggested that physical stimuli following postoperative adhesion and attachment between the iliac artery and small bowel had contributed to the formation of the fistula, although chronic infection of the graft could not be excluded completely. Despite fibrillar change in the serous membrane and muscularis propria layer of the intestine, inflammatory cell infiltration was sparse at the edge of the fistula. Moreover, there was no sign of phagocytosis, indicating no destructive infection of the graft. The arterial wall exhibited fibrous thickening, however, possibly due to continuous physical stimulation. These pathological findings and the clinical course suggested that the persistent pulsation of the aorta had physically degraded the attached intestinal wall and aortic membrane, inducing erosion of the bowel tract. Fatal bleeding appears to have occurred when the vulnerable inner membrane of the attached aorta finally ruptured.

In terms of the underlying mechanism of fistula formation, the present findings reaffirm the importance of avoiding adhesion between the intestinal wall and the anastomosis of the graft due to insufficient isolation of the latter [26]. Interposing native tissue between the aorta and the intestinal tract at primary surgery has been reported as effective in preventing SAEF associated with physical stimuli [1]. In the present case, the artery and intestinal tract had adhered tightly at the site of the fistula, even though the artificial graft had been covered with a longitudinally split aneurysmal wall and retroperitoneum during primary surgery. This particular fistula involved both the ileum and the distal suture line of the graft, a relatively rare site for an SAEF. The ileum might have become trapped by adhering to the closing line of the retroperitoneum after primary surgery, resulting in tight contact with the distal anastomosis of the graft. This suggests that keeping a graft isolated from the intestinal wall in the usual fashion during surgery may only be effective up to a certain point, and that further adjuvant maneuvering may be necessary to maintain postoperative isolation. Adhesive barriers could be applied to cover the closing line of the retroperitoneum.

Clinically, SAEFs present in a variety of ways, which makes a prompt diagnosis challenging [3, 8]. Whereas gastrointestinal bleeding occurs in most such cases, sepsis occurs in only approximately half on initial presentation [5, 8, 27], which is possibly because such symptoms are unlikely if the fistula was formed mainly due to mechanical stimuli. Infection might occur later, however, through contact with intestinal bacteria. Therefore, an SAEF should be suspected in cases of bloody stool or hematemesis after AAA reconstruction, even in the absence of septic symptoms. If an SAEF is the suspected cause of bloody stool, immediate exploratory surgery should be considered to prevent potentially catastrophic developments, even if the general status of the patient is stable.

Conclusions

This case report described gastrointestinal bleeding due to an SAEF. Physical stimuli were associated with its formation due to adhesion between the aorta and the intestinal wall after AAA reconstruction, indicating the importance of keeping a reconstructed aorta isolated from the intestine. Clinicians should suspect SAEF in patients with bloody stool after aneurysm surgery.