Clinical features and diagnosis of abdominal aortic graft or endograft infection

INTRODUCTION — Infection involving an aortic prosthesis is a challenging clinical problem for the vascular surgeon. A prosthetic graft may have been placed using an endovascular (ie, placement of an aortic endograft) or open surgical approach (ie, placement of an aortic graft) for a variety of indications.

Endovascular aortic repair has become the preferred approach for treating aortic aneurysms [1-3]. Traditional open aortic aneurysm repair is generally reserved for juxtarenal aortic aneurysms not amenable to an endovascular approach mainly due to anatomic reasons. However, advanced endografts are increasingly used to treat aortas with challenging anatomy such as short necks, allowing treatment of juxtarenal aneurysms, and even those involving the visceral arteries [4]. Aortoiliac/femoral bypass grafting continues to be used for aortoiliac occlusive disease, although more typically after an endovascular option was either not possible or after all endovascular options were exhausted.

The clinical features and diagnosis of aortic graft infection or aortic endograft infection limited to the abdominal aorta and iliofemoral arteries, whether placed using traditional open or endovascular grafting techniques, are reviewed. The clinical features and diagnostic issues are similar; however, differences, where they exist, are noted below.

TIMING AND INCIDENCE

Early versus late aortic graft/endograft infection — Aortic graft infection (AGI) or aortic endograft infection (AEGI) can present early following placement or later during follow-up. Differentiating between early and late AGIs has clinical significance; early graft infection is indicative of entire graft involvement.

●Early infection – Early AGI or AEGI is defined as infection within the first four months after graft placement [5].

●Late infection – Late AGI or AEGI is defined as infection occurring later than four months to years after graft placement [5-7].

Aortic graft infection — The incidence of AGI is estimated to be between 0.5 and 3 percent. Most early AGIs are related to a graft that became contaminated in the perioperative period or a graft was placed into an infected field. Early graft infection occurs during the period when the graft has not fully incorporated into the adjacent tissues [6,8,9]. A newly placed aortic graft does not fully incorporate into the surrounding tissues for several months. Thus, early graft infections end up involving the entire graft.

The graft configuration affects the incidence of AGI. Aortic tube grafts and aortoiliac bifurcated grafts are often used for aneurysmal disease, whereas aortobifemoral grafts are more often used for aortoiliac occlusive disease. Extension of the graft into the femoral region increases the incidence to as much as 3 percent, with an associated mortality as high as 50 percent [10-12]. (See "Open surgical repair of abdominal aortic aneurysm", section on 'Aneurysm extent'.)

The type of material used also affects the incidence of AGI [8,11-14]. Differences in graft material type and construction may be a determinant of graft vulnerability to infection. In one experimental study, bacterial adherence was higher for polyester (ie, Dacron) compared with expanded polytetrafluoroethylene [15]. Most traditional open aortic grafts, particularly those for aortoiliac or aortofemoral reconstructions, are made of polyester. (See "Open surgical repair of abdominal aortic aneurysm", section on 'Grafts for open aortic repair'.)

Aortic endograft infection — The incidence of AEGI has been reported to be <1 percent (range 0.3 to 2 percent) [11,16]. The true incidence may be higher due to unrecognized and therefore underreported infectious complications of aortic endograft repair. The incidence of AEGI will likely increase with the increasing number of endovascular interventions being performed. Like open aortic grafts, endografts that closely appose the vessel wall can also become incorporated. However, with aneurysm repair, there is little wall apposition in the aortic segment.

In most series, AEGI has been reported within the first two years of the initial procedure [1,17-19]. In a review of 36 patients, the average time from the initial surgery was 589 days (range 43 to 2466 days) [17].

In some cases, placement of an aortic endograft is associated with a periprocedural inflammatory response with leukocytosis and low-grade fever after implantation that does not represent a graft infection. (See "Endovascular repair of abdominal aortic aneurysm", section on 'Prophylaxis for postimplantation syndrome' and "Complications of endovascular abdominal aortic repair", section on 'Postimplantation syndrome'.)

MICROBIOLOGY — It is crucial to obtain accurately identify the organism(s) involved in aortic graft infection (AGI). This includes initial blood cultures, cultures from any fluid collections, as well as tissue or graft cultures obtained in the operating room at the time of graft removal. (See 'Blood cultures' below and 'Obtaining fluid for culture' below.)

The number or organisms and particular species has a direct impact on survival. It also affects surgical planning and can help shed light on the etiology of AGI [11,20,21].

●The most common species identified is staphylococcal species.

●When blood cultures are also positive, infection may be invasive, with possibly the entire graft infected.

●Aortoenteric erosion or aortoenteric fistula are suspected if cultures yield gram-negative, anaerobic, or polymicrobial organisms, especially when there is extraluminal gas around the duodenum on computed tomographic (CT) angiography.

●Anaerobic micro-organisms, such as Bacteroides fragilis with positive blood culture, have a higher probability of aortoenteric erosion or aortoenteric fistula.

●Fungal culture results can take longer to return. The local microbiology laboratory may not be able to plate complicated fungal infections.

There are few differences comparing microbiology for aortic grafts placed with open surgery and aortic endografts. Like the staphylococcal species, low-virulence organisms have fared better than gram-negative micro-organisms, similar to trends observed in open AGI series [19,20,22].

Outcomes are better for cultures that yield a single organism. Polymicrobial graft infections tend to do worse, with gram-negative and fungal infections doing very poorly [20]. In one series, five-year survival was significantly less in aortic endograft infection with gram-negative bacteria, 56 versus 31 percent [19].

While it is essential to obtain accurate culture data, the surgeon should avoid reacting to cultures considered contaminants. Swabs of superficial groin wounds should be avoided. Deep tissue and/or aspirated abscess fluid cultures are more reliable. If fluid aspirate is obtained, the addition of a portion of this to blood culture bottles is recommended. If cultures have growing gram-negative or anaerobic growth, they are considered accurate [8,12,13]. Mechanical grinding or sonication techniques have been used to improve the yield for grafts infected with biofilm-producing bacteria such as Staphylococcus epidermidis [14].

Involving infectious disease specialists sooner than later is recommended.

ETIOLOGY — The mechanism responsible for graft infection most likely occurs in the perioperative period or soon thereafter. Mechanisms include graft contamination during surgery, bacteremic seeding of the graft from a remote infection, or contiguous infection in the vicinity of the prosthesis [10]. One or more mechanisms may be responsible. As an example, bacteremia from a urinary tract infection caused by urinary obstruction with hydronephrosis can infect an aortic graft via the bloodstream as well as via contiguous contamination.

Graft contamination — Graft contamination is caused by a break in sterile technique (eg, lack of adherence to control measures, graft contacting the skin). For surgically placed aortic grafts, other examples include inadvertent bowel, biliary, or urinary tract injury during open surgery. Endografts can also be contaminated at the time of placement. One series reported that 14 percent or endografts were contaminated; however, the authors did not elaborate on the exact mechanism [19]. Such graft infections can manifest in the perioperative period but can also not be apparent until many months after the index procedure.

Another mechanism of direct graft contamination is the placement of a new aortic graft or endograft in an infected field, such as repairing a mycotic aneurysm or covering an aortic ulcer that may already be infected. These grafts are considered infected at the outset, and if the patient's medical condition allows, they should undergo definitive repair, ideally using autogenous conduit placed with an open procedure. For frail, very high-risk patients, endovascular aortic repair may suffice as the definitive procedure, but more so for palliative reasons. (See "Overview of infected (mycotic) arterial aneurysm", section on 'Role of endovascular techniques'.)

Bacteremic seeding — The risk of aortic graft or endograft infection associated with the development of bacteremia differs depending on the virulence of the bacteria, the patient's immunologic status, and the nature of the graft. Unfortunately, there is no good way to quantify the risk.

In the early period, the graft is unincorporated and vulnerable to bacteremia [23]. Over time, as a graft becomes incorporated, it is less vulnerable but can still become infected. The quantity of bacteria in the bloodstream required to infect a vascular prosthesis is unknown. Conceivably, a lower bacterial count can infect a newly placed graft or a graft in a patient who is immunocompromised [8,11,12].

Once bacteria adhere to the graft, they produce glycocalyx material that induces inflammatory cytokines [13,14]. With time, the infection and inflammatory response can extend into the perigraft tissues.

Remote infection at the time of graft placement — Remote infection (eg, urinary tract infection, respiratory infection) can seed and potentially infect a newly placed aortic graft or endograft. The bulk of the literature discussing remote infection relates to open aortic graft surgery; however, placement of any vascular prosthesis should be avoided in patients identified with remote infection because of the increased risk for graft infection [10-12,24]. Patients with poor dental hygiene or an active tooth or periodontal infection are also at a higher risk and can easily be overlooked [19,24]. The patient should be completely treated before placement of any prosthetic material whenever possible. However, this may not be feasible under emergency clinical scenarios [24]. (See 'Preventing graft infection' below.)

Despite such recommendations, aortic graft placement, particularly endograft placement, continues to occur when remote infection is present [11,19]. In a large series of aortic endograft infections (AEGIs), one-third of patients had a remote infection at the time of initial graft placement [19].

Postoperative ischemia after aortic surgery can result in bacterial translocation. It is not uncommon for the sigmoid colon to be ischemic after an emergency aortic procedure, which can put a newly placed graft at a higher risk for infection, regardless of whether there is only mucosal sloughing or full-thickness necrosis. (See "Colonic ischemia".)

Remote infection following graft placement — Although a well-incorporated graft is less susceptible, bacteremia from a remote infection (eg, periodontal infection, urinary tract infection, respiratory infection) can still infect an aortic graft [8,11]. Endografts are susceptible to infection after initial insertion, especially within the first couple of months [18,19,25]. Grafts that are subjected to graft trauma such as repeat catheterizations or multiple thrombectomies are susceptible to bacteremic seeding.

Instrumentation following graft placement

●Vascular – It is hypothesized that vascular instrumentation increases the risk for infection. Iatrogenic injury during catheterization of the aorta to evaluate an aortic graft or endograft, to perform thrombectomy, or for other procedures (eg, heart catheterization) can render aortic grafts more susceptible to infection. For endografts, an additional risk factor is the presence of an endoleak. In the largest series reported, 25 percent of the patients with AEGI had an endoleak with initial graft placement, with nearly half of those patients undergoing some procedure to treat endoleak, including direct aortic sac instrumentation [18,19,25]. It is plausible that the instrumentation for treatment of an endoleak may have led to contamination.

●Other – Bacteremia induced by instrumentation of the gastrointestinal or the genitourinary tract also increases the risk of graft infection, particularly in the first months following graft placement [18,19,25]. Procedures such as cholecystectomy or bowel resection will invariably introduce bacteria in the bloodstream. These procedures should be avoided in the postoperative period for four to six weeks, whenever possible. (See 'Preventing graft infection' below.)

Surgical site infection and other contiguous infections — Surgical site infection (SSI) is a known risk for aortic graft infection [5,10,12,26]. Antimicrobial prophylaxis prior to graft placement (with redosing during surgery as needed) is a critical measure aimed at preventing vascular graft infection (table 1). In addition, poor surgical technique, undermining skin flaps, postoperative hematoma, or lymphocele can all lead to SSI, which in turn can infect a newly placed graft. (See 'Blood cultures' below and "Overview of control measures for prevention of surgical site infection in adults" and "Overview of the evaluation and management of surgical site infection".)

An exposed graft is considered a vascular emergency and must be managed urgently. If a single groin is involved, the infection may be localized to the ipsilateral limb. If both groins are involved, then the entire graft is very likely infected [27]. Since most endovascular grafts are performed percutaneously, SSI in the groin area is rare.

Other contiguous infections can also contaminate an aortic graft. Diverticular abscess or perforated viscus can directly infect an aortic graft. Movement of the graft adjacent bowel can lead to aortoenteric erosion, which can contaminate the graft and progress to become an aortoenteric fistula. Spinal hardware infection, disc infection (particularly in the lower lumbar disc space), and abscess in the psoas muscle are other potential risk factors. Similarly, the ureters traverse the retroperitoneum near the iliac arteries. Hydronephrosis with urinary tract infection has a high likelihood of contaminating the limbs of an aortofemoral graft due to contiguous infection; bacteremia can also contribute.

RISK FACTORS — Patients who are immunocompromised (eg, posttransplant, malignancy undergoing chemotherapy, chronic corticosteroid therapy, human immunodeficiency virus) have a much higher risk for developing infectious complications. Other medical comorbidities such as cirrhosis, end-stage kidney disease, uncontrolled diabetes mellitus, and malnutrition all contribute to lowering host defenses and increase susceptibility to graft infection [11,12,28].

Grafts placed urgently or under emergency circumstances (eg, ruptured aneurysm, mycotic aneurysm) have a higher risk for developing a graft infection [12,29,30]. The immunocompromised patient or patient with a pre-existing infection is at a higher risk of infectious complications when undergoing emergency procedures. During emergency aortic surgery, the risk of colonic ischemia is also higher. While the colon may not show frank necrosis, bowel ischemia can result in bacterial translocation and bacteremia. (See "Colonic ischemia".)

CLINICAL PRESENTATIONS — The clinical presentation of aortic graft infection (AGI) varies widely from minimally symptomatic in patients with low-virulence organisms to hemodynamic collapse in patients with overt sepsis, aortic rupture, or aortoenteric fistula. There are some notable differences between AGI and aortic endograft infection (AEGI).

Patients can present with systemic or localized symptoms, alone or in combination. Localized symptoms and signs of graft infection can manifest as abdominal pain or as surgical site, limb, or gastrointestinal events.

●In a review of 187 AGIs, clinical presentations included an open groin sinus/wound (43 percent), femoral pseudoaneurysm (36 percent), constitutional symptoms (32 percent), limb ischemia (29 percent), sepsis (21 percent), aortoenteric erosion/fistula (14 percent), and bleeding (12 percent) [20].

●Most patients with AEGI (70 percent) present with pain, fever, and leukocytosis, while approximately 33 percent present with weight loss, fatigue, or generalized weakness [31]. In a review of 36 endograft infections, clinical features included leukocytosis (63 percent), fever (56 percent), and pain (58 percent), which were present an average of 65 days (range 0 to 514 days) before explantation [17].

Systemic symptoms — Symptoms related to cavitary infection are usually systemic. Systemic symptoms are more common with AEGI because the infection is contained within the confines of the aortic sac, and unlike standard aortic grafts, which may have a well incorporated uninvolved segment, the entire endograft is usually infected.

●Sepsis – Sepsis represents advanced infection and is common with early graft infection, more virulent organisms, and infections that affect the entire graft.

●Failure to thrive – Symptoms are often nonspecific and can variably include generalized malaise, fatigue, low-grade fever, night sweats, loss of appetite, and anemia (eg, with graft-enteric erosion). Fever and/or night sweats occur when the entire graft is involved [11,12,30].

Localized symptoms

Abdominal pain — A common presenting symptom is poorly defined abdominal/flank pain. Almost two-thirds of the patients will have either abdominal/flank pain or low-grade fever, or both. Abdominal pain and groin pain (aortofemoral grafts) may or may not be associated with proximal or distal anastomotic pseudoaneurysm formation, respectively.

Surgical site events — (See "Overview of the evaluation and management of surgical site infection".)

●Surgical site infection – In the case of an aortofemoral graft, the area surrounding the incision may be swollen and warm with cellulitis, with or without skin breakdown or evidence of a draining sinus tract. Lymph leakage (lymphocele) or seroma formation postoperatively increases the risk of graft infection. Any delay in wound healing or groin swelling in the early postoperative period may represent a graft infection. If a graft becomes visible in a groin wound, it is considered infected and must be dealt with urgently.

●Distal anastomotic disruption – Hemorrhage or pseudoaneurysm formation at the distal anastomosis of an aortofemoral bypass graft can be from invasive infection of the graft at the suture line. Anastomotic bleeding in the perioperative period related to early graft infection can be easily mistaken for technical complications. Within 7 to 10 days of surgery, anastomotic bleeding should be presumed to be from an invasive infection and addressed urgently.

Acute limb events — Ischemia due to graft limb thrombosis or embolism can be a sign of graft infection and is usually discovered during surgery for thrombectomy when the graft is found to be unincorporated. Evidence of any septic emboli in any distal vascular tree is an indication for early intervention. (See "Clinical features and diagnosis of acute lower extremity ischemia".)

Gastrointestinal events — Gastrointestinal bleeding can be a sign of an aortoenteric erosion or a herald sign of an aortoenteric fistula. Gastrointestinal bleeding can range from occult anemia to massive, life-threatening hemorrhage. Any communication of the graft with the gastrointestinal tract is regarded as contaminating the aortic graft or endograft and must be managed accordingly. (See "Aortoenteric fistula: Recognition and management".)

●During open aortic surgery, the graft is typically covered using the aneurysm sac, and the retroperitoneal tissues are reapproximated. In some cases, additional omentum can be used to create a barrier between the aortic graft and surrounding structures including the duodenum and small bowel. Nevertheless, an aortoenteric fistula can occur and was the presentation in 14 percent in one review [20].

●For aortic endografts, the overlying aorta/aneurysm sac should theoretically protect the graft from the bowel; however, almost a fourth of the patients (25 percent) with endograft infection present with aortoduodenal fistula or rupture [17,19]. It is plausible in some of these cases that a primary aortoenteric erosion was present but unrecognized prior to endograft placement.

Abnormalities on surveillance imaging — Following aortic grafting, the frequency of surveillance imaging varies according to the type of repair, being more frequent for aortic endografting to evaluate for endoleak and any signs of sac expansion. For aortic grafts, imaging is less frequent, but some patients may present with concerning features having attributed any symptoms to another cause.

●For endografts, a rapid increase in the diameter of the aneurysm sac might be seen with no obvious endoleak on CT angiography.

●For aortic tube grafts or aortoiliac/femoral grafts, the development of pseudoaneurysm is a sign of possible graft infection.

DIAGNOSTIC EVALUATION — The clinician must maintain a high index of suspicion for a diagnosis of aortic graft infection (AGI) or aortic endograft infection (AEGI) given the broad range of clinical presentations.

The objective of the diagnostic evaluation is to not only to judge the likelihood of graft infection but also to determine the severity of infection, the microbiology (blood culture, fluid aspiration for culture), and the overall medical condition of the patient. It is crucial to determine if there is clinical evidence graft-enteric communication; however, this may only be possible with abdominal exploration.

The diagnosis of graft infection (see 'Diagnosis' below) relies heavily on demonstrating features consistent with infection on imaging. For most patients, whenever AGI is suspected, we initially obtain CT angiography, if possible. Complementary imaging modalities may be needed to support a suspected diagnosis. For patients with emergency presentations, the diagnosis may be established in the operating room.

Laboratory examination — Basic laboratory studies include complete blood count, electrolytes, renal function, and coagulation profile. Inflammatory markers (eg, erythrocyte sedimentation rate, C-reactive protein, procalcitonin) are also obtained. The stool should also be tested for blood. In patients with an aortic graft, occult blood in the stool with or without anemia is considered a graft-enteric communication until proven otherwise.

Other important laboratory studies include assessing protein levels to evaluate nutrition status. (See "Clinical assessment and monitoring of nutrition support in adult surgical patients", section on 'Initial assessment of nutritional status'.)

Inflammatory markers — Patients with AGI may or may not have leukocytosis on the initial presentation [32]. Patients who do present with leukocytosis and/or fever may have an invasive infection.

Erythrocyte sedimentation rate, C-reactive protein, and procalcitonin levels serve as markers for inflammation and can be drawn at baseline and then trended to assess the severity of infection at the initial presentation and the subsequent response to treatment [28,33].

Blood cultures — Blood cultures should be obtained in any patient suspected to have a vascular graft, particularly those who exhibit systemic symptoms. It is recommended that the patient is off antibiotics for 24 to 48 hours, if clinically possible, before blood cultures are drawn, with one set preferably from an arterial source. (See "Detection of bacteremia: Blood cultures and other diagnostic tests".)

Cultures are negative or not available in 17 to 25 percent of patients [6,10,11,20,29]. This might be attributed to patients have been receiving antibiotics to treat other presumed illnesses before presenting to the surgeon, or inadequate culture techniques. Even when blood cultures drawn before antibiotic therapy, cultures yield positive results in only 25 to 40 percent of patients [8,11,12,17,19,25].

Endoscopy — Gastrointestinal bleeding in a patient with an aortic graft or endograft, either in the form of gross bleeding (hematemesis or hematochezia) or as hemoccult positive stool sample, supports the decision to perform endoscopy. The role of endoscopy is primarily to rule out other etiologies for gastrointestinal bleeding. If none are identified, then bleeding that occurs in a patient with an aortic endograft is presumed to be from an aortoenteric fistula. (See "Aortoenteric fistula: Recognition and management".)

Endoscopy should include evaluation of the third and fourth portions of the duodenum. The observation of graft material on upper GI endoscopy is considered evidence of graft infection, though this finding is uncommon even in the setting of known aortic graft-enteric erosion or aortoenteric fistula. Negative upper endoscopy does not rule out the possibility of graft-enteric erosion or fistula.

Vascular imaging — CT angiography is typically the initial imaging study obtained under most clinical circumstances when graft infection is suspected. However, in the first couple of months following aortic grafting, no imaging study can reliably distinguish between a graft infection and postsurgical changes. Certain imaging abnormalities identified beyond eight weeks are considered abnormal and may represent infection. The advantages and disadvantages of the various imaging modalities and their accuracy are summarized in the table (table 2).

CT angiography — CT angiography is the most frequently initial obtained imaging modality to evaluate for graft infection (table 2) [34,35].

CT angiography provides extensive details about arterial anatomy, including the status of the inflow and outflow vessels, the presence of other vascular pathology (eg, atherosclerosis), and high-definition images of the soft-tissue changes that may be diagnostic of infection. These include perigraft fluid, air around the graft, abscess formation (image 1), pseudoaneurysm, adjacent bowel thickening (image 2), and adjacent stranding from inflammation [7,36,37].

For the diagnosis of late graft infection, CT angiography has a positive predictive value near 90 percent and is often the only study needed to diagnose and plan treatment. If the diagnosis is confirmed (see 'Diagnosis' below), further imaging is generally not necessary.

For early graft infections, CT angiography can be inconclusive.

●AGI – For early AGI, especially within the first six to eight weeks, it may be difficult to distinguish between postsurgical changes and graft infection. After four weeks, there should not be any air seen around the graft, and certainly none beyond six weeks.

●AEGI – Air is commonly observed within the aneurysm sac immediately following endograft insertion. This finding typically resolves spontaneously within the first few weeks after implantation and does not represent a pathological condition, thus does not deserve further investigation. However, air within the aneurysm sac beyond six to eight weeks after implantation is considered abnormal and may be due to infection. The presence of metal in endograft stents can cause artifacts that may be misleading and interpreted as air.

Magnetic resonance angiography — Magnetic resonance (MR) imaging is at least as sensitive as CT angiography for diagnosing AGI. In a review of 59 patients with suspected AGI, the positive predictive value was 95 percent, and a negative predictive value of 80 percent [38]. Nevertheless, MR imaging has not been as used as frequently, often for practical reasons (table 2), or because the study is contraindicated [39]. A diagnosis of AGI with MR imaging is more typically incidental when the study is obtained to investigate complaints not initially thought to be related (eg, spine MR for back pain).

MR has similar drawbacks as CT angiography and cannot reliably distinguish between early postoperative changes versus infection. An advantage of MR imaging is that it allows tissue characterization; the T-1 weighted images may show fluid in AGI with low to medium signal intensity [9,40]; high signal intensity can be seen on T-2 weighted images [9,36,40].

With aortic endografts, metal stents can be misleading with artifact and image dropouts. Although most endografts are considered MR imaging compatible, some manufacturers’ instructions include warnings about obtaining MR imaging too soon after implantation.

Duplex ultrasound — Duplex ultrasound (DUS) is a noninvasive, low-cost imaging test that can be used as an initial screening tool if AGI is suspected (table 2).

For patients with grafts that extend to the groin (eg, aortofemoral graft), DUS is a useful adjunct for the initial evaluation, particularly for the limbs of the graft. DUS can detect and moderate sized fluid collections or abscesses and can identify the presence of fluid around the graft as the "halo sign" [41]. For patients presenting with limb ischemia from suspected graft limb thrombosis or embolization, the primary use for DUS is to assess arterial flow.

DUS has limited value for diagnosing AEGI and any positive results require confirmation. The diagnosis of AEGI is never made solely based on DUS. However, there is always a value of detecting perigraft fluid around the iliac limbs of an endograft in questionable cases with no overt CT angiography findings. However, it requires a high degree of technical expertise to locate minimal fluid around an endograft. It is most useful 10 to 12 weeks after the initial procedure.

Other diagnostic studies — When CT angiography is equivocal, other diagnostic methods (table 2) may be relied upon in specific circumstances, especially in the early postoperative period in patients with vague symptoms. Depending on the clinical circumstances, these may include obtaining fluid for culture, nuclear studies, fluorodeoxyglucose positron emission tomography (FDG-PET), and sinography.

Obtaining fluid for culture — CT- or DUS-guided aspiration of any fluid collections can provide accurate microbiology information and confirm the diagnosis [8,10,42]. (See 'Diagnosis' below.)

For aortic endografts, the aneurysm sac can be entered with direct visualization of the endograft through laparoscopy. It is an appealing option for the surgeon, as there are minimal to no adhesions, and the aortic aneurysm sac is approachable. The fluid aspirated has a high yield for positive cultures. It may be an option when endograft preservation is selected; however, this technique is not widely practiced, as most surgeons opt for CT-guided aspiration. [43].

FDG-PET/CT — FDG-PET can also be used to detect graft infection (table 2) [36,44-46]. It useful for patients with equivocal CT imaging and in patients with vague symptoms, particularly in the first four to six weeks after the initial procedure. The differentiation of infection from inflammation can be seen in the pattern of glucose uptake. Inflammation may show diffuse uptake, whereas infection lights up as focal uptake [37,46]. While FDG-PET does not provide anatomic information, it is highly accurate when combined with CT angiography and can provide high-quality anatomic details [36]. PET-CT angiography has shown promising results with a high positive predictive value and even higher negative predictive value compared with CT angiography alone [27,45,47].

Nuclear white blood cell scanning — Nuclear leukocyte scintigraphy is a useful adjunct for the diagnosis of AGI when the diagnosis is in question (table 2).

The two main techniques are indium-111-labeled leukocyte immunoglobulin-G scan and the gallium-67 scan. Indium-111 has an advantage over gallium-67, as it has minimal bowel uptake. The indium-111 is 100 percent sensitive but can be falsely positive in the early postoperative period. A positive nuclear scan in late graft infections is diagnostic, but a negative scan does not rule out a graft infection.

Sinography — Sinography has been used to show communication between a suspected surgical site infection and a deeper graft to differentiate superficial from deep infections. The contrast is injected into the sinus tract to determine the tract's depth and whether it communicates with a perigraft cavity or the graft. With the availability of other sensitive diagnostic modalities, sinography is not as commonly used.

DIAGNOSIS — Aortic graft infection (AGI) or endograft infection is diagnosed based on a combination of the clinical presentation and findings on the diagnostic evaluation including supporting evidence from laboratory tests, cultures (blood culture, fluid culture), and imaging studies. Confirmation of the diagnosis relies heavily on aortic imaging demonstrating features consistent with periaortic infection with or without evidence of bacteremia. (See 'Diagnostic evaluation' above.)

The diagnosis may be obvious, such as purulence from a groin incision with an exposed graft, imaging findings of an abscess around the graft or the presence of pseudoaneurysm, or sepsis and no other identified source in a patient with an aortic graft. Alternatively, the patient can present as a clinical dilemma with vague, nonspecific symptoms, unexplained leukocytosis, elevated inflammatory markers, and no supporting radiologic evidence [4,5,12,30,31].

The diagnostic criteria provided below may be helpful for determining the likelihood of aortic graft or endograft infection. (See 'MAGIC diagnostic criteria for aortic graft infection' below.)

MAGIC diagnostic criteria for aortic graft infection — The Management of AGI Collaboration (MAGIC) introduced criteria to help establish the diagnosis of an aortic graft/endograft infection based on "major" or "minor" clinical, imaging, and laboratory findings [48].

●Vascular graft/endograft infection is suspected if a single major criterion or two or more minor criteria from different categories are present.

●Vascular graft or endograft infection is diagnosed when there is one major plus any criterion (major or minor) from another category.

The MAGIC criteria were validated in a prospective cohort study of patients with definite and suspected graft infection [34]. The Vascular Graft Cohort (VASGRA) cohort categorized patients as "diseased" or "not diseased." Using the MAGIC criteria, the calculated sensitivity for MAGIC was 99 percent (95% CI 96-100) and specificity was 61 percent (95% CI 52-70).

Major criteria

●Clinical/surgical – Pus around graft or in aneurysm sac at surgery; open wound with exposed graft or communicating sinus; aortoenteric fistula development; graft that was placed in an infected field.

●Radiologic/imaging – Perigraft fluid on CT ≥3 months after graft placement; perigraft gas on CT ≥7 months after graft placement; increase in perigraft gas on serial imaging.

●Laboratory – Organisms recovered from an explanted graft; organisms recovered from intraoperative specimen; organisms recovered from a percutaneous, radiologically guided aspirate of perigraft fluid.

Minor criteria

●Clinical/surgical – Localized clinical features of graft infection; fever >38°C with graft infection the most likely cause (other causes ruled out).

●Radiologic/imaging – Other suspicious perigraft gas/fluid/soft tissue inflammation; aneurysm expansion; pseudoaneurysm formation; focal bowel wall thickening; discitis/osteomyelitis; suspicious metabolic activity on fluorodeoxyglucose positron emission tomography CT; radiolabeled leukocyte uptake.

●Laboratory – Positive blood culture with no apparent other source except graft infection; abnormally elevated inflammatory markers (white blood cell count, erythrocyte sedimentation rate, c-reactive protein test) with graft infection as the most likely cause.

Other classifications — The United States Centers for Disease control criteria for establishing surgical site infection (table 3) may be potentially helpful for considering early graft infection, but does not specifically consider vascular graft involvement. (See "Overview of the evaluation and management of surgical site infection", section on 'Diagnosis'.)

Other classifications, which do consider vascular graft involvement, do not necessarily aid in the diagnosis [26,49,50].

Differential diagnosis — Because the clinical presentation can vary so widely, the differential diagnosis for AGI is broad and includes the differential diagnosis for the following. The approach to these symptoms and signs is reviewed in the linked topics.

●Other causes of abdominal pain (See "Causes of abdominal pain in adults".)

●Other causes of upper gastrointestinal bleeding (See "Causes of upper gastrointestinal bleeding in adults" and "Etiology of lower gastrointestinal bleeding in adults".)

●Other causes of fever (See "Fever in the surgical patient".)

●Unrelated bacteremia (See "Clinical approach to Staphylococcus aureus bacteremia in adults" and "Gram-negative bacillary bacteremia in adults".)

Placement of an aortic endograft is associated with an inflammatory response with mild leukocytosis, typically above 11,000 white blood cells/microL [51]. Low-grade fever after graft implantation in the perioperative period should not be confused with a graft infection. However, imaging abnormalities beyond eight weeks are not considered normal and may represent infection. In most series, the endograft infection has been reported within the first two years of the initial procedure. Procalcitonin levels may help in distinguishing postimplantation syndrome from bacterial infection. (See "Complications of endovascular abdominal aortic repair", section on 'Postimplantation syndrome' and "Endovascular repair of abdominal aortic aneurysm", section on 'Prophylaxis for postimplantation syndrome'.)

PREVENTING GRAFT INFECTION — We use the following measures, many of which have not been rigorously evaluated, in an effort to prevent the devastating consequences of aortic graft or endograft infection [28,52].

●Preoperatively:

•Scrub the surgical site with bactericidal soap 24 hours before incision, especially in the groin area. (See "Overview of control measures for prevention of surgical site infection in adults", section on 'Skin antisepsis'.)

•Control any remote infection and delay the procedure until the infection is completely eradicated. (See "Overview of control measures for prevention of surgical site infection in adults", section on 'Remote infection'.)

•Avoid prolonged stays in the hospital while waiting for surgery.

•Encourage the patient to undergo dental examination to identify and treat any potential tooth or periodontal infection.

•Inquire and optimize any issues with prostatic hypertrophy in males.

•Improve nutrition status (prealbumin >15). (See "Clinical assessment and monitoring of nutrition support in adult surgical patients".)

•Delay elective vascular procedures for patients undergoing chemotherapy or corticosteroid therapy until after their course of treatment is completed.

●During surgery:

•Allow alcohol-based solutions to dry before placing drapes.

•Apply iodine-impregnated impermeable plastic adhesive drapes (Ioban).

•Clip hair instead of shaving it at the incision site. (See "Overview of control measures for prevention of surgical site infection in adults", section on 'Hair removal'.)

•Mark the femoral artery's location on the skin with the help of ultrasound before making groin incisions. This is especially beneficial in an obese patient to minimize thin skin flaps.

•Change gloves before handling an aortic graft at the time of anastomosis.

•Soak aortic grafts in an antibiotic solution before the implantation. While various coatings (eg, silver, antibiotics) have been tried to reduce the incidence of aortic endograft infection, in a systematic review, no significant protection was observed [53].

•Avoid graft or endograft contact with the skin.

•Redose antibiotics during surgery if the procedure takes longer than four hours. Continue antibiotics beyond 24 hours in immunocompromised patients.

●Postoperative:

•Use negative pressure wound therapy dressings on postoperative incision(s) [54]. (See "Negative pressure wound therapy", section on 'Prophylactic use'.)

•Use oxygen therapy via nasal cannula to keep tissue oxygenation levels higher than 95 percent during hospitalization [55].

•Optimize control of blood sugar. (See "Overview of control measures for prevention of surgical site infection in adults", section on 'Glucose control'.)

SUMMARY AND RECOMMENDATIONS

●Aortic graft/endograft infection – Infection involving aortic graft materials is an uncommon but challenging clinical problem. The graft may have been placed using an endovascular or open surgical approach. Aortic graft infection (AGI) or aortic endograft infection (AEGI) can present early following placement or later during follow-up. (See 'Early versus late aortic graft/endograft infection' above.)

●Etiology – The mechanisms causing AGI generally include contamination during surgery, infection via the bloodstream (remote infection), and infection in the vicinity of the prosthesis (contiguous infection); more than one mechanism can occur together. Patients who are immunocompromised and grafts placed under emergency circumstances have a higher risk for graft infection. (See 'Etiology' above.)

●Clinical features – The clinical presentation of AGI/AEGI varies widely from minimally symptomatic in patient with low-virulence organisms to hemodynamic collapse in patients with overt sepsis, rupture, or aortoenteric fistula. Patients may have systemic symptoms (eg, sepsis, failure to thrive), localized symptoms (eg, surgical site infection [SSI], acute limb ischemia, gastrointestinal bleeding), or both. An initial clinical presentation with systemic symptoms is more common with AEGI. (See 'Clinical presentations' above.)

●Diagnostic evaluation – The clinician must maintain a high index of suspicion for a diagnosis of AGI or AEGI given the broad range of clinical presentations. The objective of the diagnostic evaluation is to judge the likelihood of graft infection but also to determine severity, microbiology (blood culture, fluid aspiration for culture), and the overall medical condition of the patient. For most patients, we initially obtain CT angiography, if possible. Complementary imaging modalities (eg, duplex ultrasound [DUS], fluorodeoxyglucose positron emission tomography [FDG-PET]) may be needed to support a suspected diagnosis. For patients with emergency presentations, the diagnosis may be established in the operating room. Determining whether there is a graft-enteric communication is crucial. (See 'Diagnostic evaluation' above.)

●Diagnosis – AGI or AEGI is diagnosed on a combination of the clinical presentation and findings on the diagnostic evaluation including supporting evidence from laboratory tests, cultures (blood culture, fluid culture), and imaging studies. The diagnosis may be obvious based on the clinical features, but often relies on imaging studies demonstrating features consistent with infection. The MAGIC diagnostic criteria may be helpful for determining the likelihood of aortic graft or endograft infection. (See 'Diagnosis' above.)