DISCUSSION
Aortoenteric fistulas (AEF) constitute a rare clinical entity,
concerning a pathological connection between the aorta and the
gastrointestinal tract. Although their incidence barely reaches 4%
(between 1.6 and 4%), they are associated with high mortality rates
(24% to 45.8%).2, 3
They are divided into two major categories, the Secondary AEF (SAEF),
which most commonly occur after previous aortic surgery, either open of
endovascular, and the Primary AEF (PAEF), that take place spontaneously
without any previous intervention in the aorta and are even rarer
(incidence <0.07%)4.
High-risk patients of developing SAEFs include those who undergo
emergent surgery for a ruptured aneurysm, have post-operative
complications such as reoperation or bowel injury, and those with
endoleaks or stent migration.5
Most commonly, an AEF presents with an initial or multiple “herald
bleedings”, meaning episodes of transient, seemingly self-limiting and
minimal lower GI bleeding. A catastrophic bleed follows, resulting in
significant hemodynamic instability, requiring urgent surgical repair.
Other clinical signs and symptoms might be sepsis (44%), abdominal pain
(30%), back pain (15%), groin mass (12%), and abdominal pulsatile
mass (6%).6
Clinical symptoms are very important in diagnosis of an aortoenteric
fistula, because of the lack of high sensitivity and specificity of a
single imaging modality. In previous reports authors suggested the
diagnostic triad of pain, sepsis, and gastrointestinal bleeding,
although the septic element is not always present in a
SAEF.8
In our case, the patient presented with septic shock. This was explained
by the infected pseudoaneurysm, which was the cause of the fistula. The
previous medical history should arise the suspicion of such a diagnosis,
creating the question of the best diagnostic imaging test.
Upper endoscopy is useful to exclude other sources of bleeding, such as
peptic ulcer disease, but a normal endoscopy result never rules out AEF.
When in suspicion of a SAEF, CT scan is considered the work-horse exam,
although literature reports variable sensitivity and
specificity.8 The most important imaging finding in an
AEF is ectopic gas in the aortic lumen or in direct contiguity to the
aortic lumen. In the most obvious cases, a direct tract of gas can be
traced directly from the involved bowel loop towards the
aorta.7 However, while highly suggestive; the presence
of gas is not completely specific. For example, perigraft soft-tissue
edema, fluid, and ectopic gas may be normal CT findings immediately
after surgery.9 However, after 2-3 months,
identification of any ectopic gas should be considered a sign of
perigraft infection with the possible presence of AEF until proven
otherwise.
Other CT findings that are suggestive of both perigraft infection and
aortoenteric fistula include pseudoaneurysm, effacement of the
periaortic fat, tethering of a bowel loop immediately adjacent to the
aorta, disruption of the aortic wall or a graft or significant graft
migration.7
The most specific sign in AEF, direct extravasation of contrast from the
aortic lumen into a intestinal loop, is especially rare to identify on
CT. Similarly, the leakage of enteric contrast directly into the
periaortic space is a highly specific sign, but extremely rare. Notably,
in a series by Hagspiel the extravasation of contrast from the aorta
into the bowel lumen was present in only 11% of
cases.14
Digital Subtraction Angiography (DSA) is frequently requested for
further evaluation of gastrointestinal bleeding. However, in most
patients with bleeding in the gastrointestinal tract, DSA offers few,
pointing of to be a poor diagnostic tool.14 On the
other hand CTA is proven to be a helpful tool for identifying disruption
of the aortic wall, outline the a pseudoaneurysm, or demonstrating the
presence of an AEF by describing direct extravasation from the aorta
into the gastrointestinal lumen.10-11
The nonspecificity of these features is responsible for a considerable
overlap with a variety of other disorders, with perigraft infection
being the most important, as it can appear identical to a fistula. Other
entities which intrigue the diagnostic procedure include aortitis,
mycotic aneurysms, and perianeurysmal fibrosis, all of which can
demonstrate periaortic inflammation, fluid, or soft
tissue.7, 12, 13
Intramural gas and aortic wall thickening in a patient with positive
blood cultures should raise suspicion of the presence of infectious
aortitis. Similarly, when a patient is presented with a saccular
aneurysm of the aorta accompanied with clinical evidence of sepsis, the
surgical team should act prompt to treat the infected aortic aneurysm
before it ruptures.7
As noted in the beginning, a secondary aortoenteric fistula occurs most
commonly after an aortic aneurysm reconstruction. Whether this
complication occurs more frequently after an open or an endovascular
repair remains open to interpretation.16, 17
Considering that arterial suture line appears to be the main
predisposing factor in open repair, endovascular treatment of AAA was
initially thought to confer little to no risk of AEF. However,
persistent endoleak, with the continued presence of blood flow into the
aneurysm sac and aneurysm sac enlargement, represents a significant
complication of endovascular aneurysm repair.18
Further, degeneration of the aneurysm neck causes graft migration with
proximal endoleak and presence of pulsatile blood flow into aneurysm sac
which eventually will lead to continued aneurysm expansion and rupture
or subsequent bowel erosion and creation of AEF.19
Identification of impending or contained rupture is critical because
these patients are at risk for frank rupture but can generally benefit
from a more thorough preoperative assessment, followed by urgent
surgery.20
Regardless the initial reconstructive surgery, when a SAEF is present,
traditionally is managed by graft explantation, wide debridement of the
infected tissues, infrarenal aortic stump ligation, and extra-anatomic
revascularization with axillo-bifemoral bypass.21
This former gold-standard procedure has been doubted, as it hides the
risk of aortic stump blow-out syndrome, specifically in long term follow
up. As a result, in situ bypass grafting using homografts, allografts,
prosthetic or vein grafts was developed. 23-25
In a series published by Kakkos et al, open surgery had higher
in-hospital mortality (246/725, 33.9%) than endovascular methods (7/89,
7.1%, p<0.001). The reduced postoperative mortality after
endovascular surgery indicated that expedient haemostasis achieved with
former methods without any further insults is probably all that is
needed in this patient population with haemmorhagic shock. The
statistical difference, however, mostly disappeared during the first
18-24 months after the procedure.22
Most common causes of death in short-term post-operative period were
irreversible shock, cardiac arrest, bleeding – either from stump
blowout or homograft rupture- sepsis, MODS and GI complications
including leak from the GI repair. During long-term follow up bleeding
(recurrent AEF, stump blowout, disruption of the proximal anastomosis or
homograft rupture), sepsis MOF or coronary syndromes were noted as the
commonest causes of death.