The incidence of aortic stent graft infection is with 0.2% to 0.7% low (1). However, there is an increasing need to be vigilant for the risk of stent graft infections as there are more patients treated using the endovascular technique. An early detection of graft infections and an adequate treatment during follow-up are also important to prevent serious complications. There have been several reports on stent graft infection with its characteristic clinical presentations and computed tomography (CT) findings (2, 3, 4, 5, 6). However, even with these characteristic CT findings, there have been few cases reported of abscess formation due to graft infection without characteristic CT findings, which could lead to either misdiagnosis or delayed diagnosis. We report a case of perigraft abscess formation that showed an increased size of aneurysmal sac without any apparent evidence of causative risk factors and was developed as a late complication 2 years after endovascular aortic repair (EVAR) for an abdominal aortic aneurysm (AAA).

A 71-year-old man underwent elective EVAR due to a progressively enlarging AAA with 72 mm size and an infrarenal neck of 4 cm. A bifurcated aortic stent graft (Endurant, Medtronic, Minneapolis, MN, USA) was inserted under local anesthesia without surgical opening through the common femoral artery. The limb of the stent graft was placed in the common iliac artery and the stent graft was successfully placed without endoleak at the appropriate position. There were no procedure-related complications or fever. Prophylactic antibiotics were administered before and after EVAR. CT scans were scheduled for the first and third month during admission and follow-up and every six months thereafter. CT evaluation of the aorta comprised pre-enhanced scanning, followed by arterial-phase CT performed with bolus tracking. Delayed imaging was not routinely performed. CT datasets were obtained by using a 64-detector row scanner (slice collimation 32 × 0.6 mm, Sensation 64, Siemens Medical Solution, Forchheim, Germany) with the following parameters: section thickness 1.0 mm, reconstruction interval 1.0 mm, pitch 1.25, 120 kV and 150 mAs. For contrast enhancement, 70 mL of non-ionic contrast media (Iomeron 400, Braco, Milan, Italy) were intravenous injected with bolus tracking. There were no post-operative complications visible on an initial CT scan taken 1 month after EVAR (Fig. 1). On the 6-month follow-up CT, there were no interval changes in the size/shape of the perigraft native aorta or in the amount of the perigraft seroma. The maximal outer diameter of the aortic wall was 70 mm and the thickness of perigraft fluid was 35 mm. The patient was afebrile and showed white blood cells (WBCs) and an erythrocyte sedimentation rate within normal ranges. The subject was admitted to our hospital due to progressive and severe abdominal pain at the follow-up after two years. At the time of admission the body temperature was 37.0℃. Blood laboratory tests revealed a WBC of 12000/µL, a segmented neutrophil count of 89%, and C-reactive protein at 22.1 mg/dL. The patient never received therapeutic or diagnostic procedures where the field of the stent graft could be traversed. The diameter of the native aneurysmal sac was increased from 70 to 80 mm and the thickness of perigraft fluid increased to a range between 35 and 42 mm on the CT scans. The CT density of the perigraft soft tissue decreased from 47 Hounsfield units (HU) to 30 HU if compared with those of the last follow-up CT scans which were taken 1 year prior to this presentation. Also were noted new radiologic findings of a thickened native aneurysmal aortic wall with slightly irregular or interrupted wall enhancement, a small eccentric luminal bulging and surrounding streaky opacities suggesting an inflammatory infiltration into the mesenteric fat (Fig. 2). The CT densities of the native aneurismal wall were 26 HU on the pre-contrast image and 44 HU on the arterial phase image. A treatment was started with intravenous administered ceftriaxone and metronidazole. Two days later, an increased diameter of the aneurysmal sac ranging from 80 to 89 mm, an increased thickness of perigraft fluid ranging from 42 to 48 mm, an advanced extent of inflammatory infiltration and a newly eccentric bulging contour of the aortic wall suggestive of impending rupture were shown in the CT scans (Fig. 3). The patient underwent a removal of the aortic stent graft along with a resection of the abdominal aneurysm which was followed by a bilateral aortoiliac bypass surgery. During surgery, there were severe inflammatory and gangrenous changes in the aneurysm sac, an abscess formation and a fluid collection between the graft and the native aneurysmal aortic wall. Large amounts of pus were evacuated and saline irrigation was done. The total amount of drained pus was approximately 1 liter. There was no evidence of an aortoenteric fistula. Pus cultures were negative and gram-staining of the pus showed gram-negative bacilli. Final results of cultures from the specimens of fluid, thrombus, aortic wall and the graft itself demonstrated no growth. The patient was medicated with intravenous antibiotics, including meropenem, vacomycin, and piperacillin/tazovactam for 2 weeks and was discharged without any specific events 2 weeks after surgery.

We present an unusual case of perigraft abscess as a late complication of EVAR that was detected on CT without any associated causative lesions during the 2-year follow-up check-up.

Late aortic stent graft infection as an increasing incidence of midterm and long-term complications after EVAR of AAA has received far less attention than other complications such as endoleak, stent graft migration and stent graft rupture. As a result, little is known about the general features and potential risk factors of aortoiliac stent graft infection (7, 8).

According to previous reports, aortic stent graft infections have been presented in roughly one-third of patients with the evidence of an aorto-enteric fistula, in one third with nonspecific signs of low grade sepsis and in the remainder with evidence of severe systemic sepsis. However, patients' symptoms were vague and nonspecific, which leads to incorrect or delayed diagnosis of stent graft infection in many cases (1). Vogel et al. (8) also concluded that aortic graft infection is associated with periprocedural infections for both endovascular and open AAA repairs. The occurrence of late graft infection was significantly higher in patients who had blood stream septicemia and surgical site infection during a periprocedural hospitalization. The patient in our case was symptom-free for 2 years and no definite causative risk factors such as periprocedural nosocomial infection, septicemia or surgical site infection were found. As we could not exclude the possibility of a low virulent periprocedural infection, further studies will be needed to achieve a more comprehensive diagnosis.

Although Staphylococcus aureus is the single most common causative organism in endograft infections, only 55% were cultured from the endograft and blood specimen (1, 4, 7, 9). In our case, blood and pus cultures as well as final cultures for fluid, thrombus, aortic wall and the graft itself showed no growth after the surgery. We assume that these results might be related to a low-grade virulent periprocedural infection in addition to the prior usage of antibiotics.

As CT findings in the acute stage of the presented case were founda diameter increase of the native aneurysmal sac with slightly thickened and irregular wall enhancement, a small eccentric luminal bulging of the native aneurysmal wall and a slight streaky opacity into the mesenteric fat. However, there were no definite CT findings of the infected aorta, such as perigraft air bubbles or severe perigraft inflammatory changes. Ferrar et al. (10) reported a low-grade infection may take a number of years to manifest as persisting endotension and aneurysm expansion and the diagnosis of endoluminal stent-graft infection should be considered in all cases of unexplained endotension. As shown in our case, even though perigraft air was not seen, radiologists should suspect abscess formation and liquefaction on the basis of clinical symptoms and unexplained endotension as well as of the basis of a decrease in the CT density of perigraft soft tissue or perigraft fluid.

In summary, we presented the rare case of a 71-year-old man who was diagnosed with stent graft infection that led to unusual perigraft abscess formation 2 years after EVAR of AAA without any definitive causative risk factors. Although clinical presentation and CT findings of aortic stent graft infection may be nonspecific, stent graft infection should be considered in case of unexplained endotension even without specific etiologies during CT follow-up.