Annals of Vascular Surgery
Volume 24, Issue 4 , Pages 554.e1-554.e5, May 2010

The Management of Aortic Stent-Graft Infection: Endograft Removal Versus Conservative Treatment

Department of Angiology and Vascular Surgery, Hospital Universitari de Bellvitge, Barcelona, Spain

published online 25 January 2010.

Article Outline

Background

Aortic stent-graft infections (ASGIs) are associated with significant mortality. We report our experience of two cases of ASGI treated differently and successfully.

Methods

Two patients presented with constitutional symptoms some months after scheduled endovascular repair of aortic aneurysm (EVAR). Patient 1 had an abscess formation around the endograft in continuity with the right groin. Due to patient comorbidities, a conservative treatment was performed. Patient 2 had an abscess formation with air surrounding the stent graft. The patient was treated successfully by endograft removal.

Results

Computed tomographic scan follow-up at 6 months from surgery showed no evidence of recurrent infection.

Conclusion

Despite the recommended treatment of ASGI being surgery, conservative treatment can be performed successfully in patients with high surgical risk, avoiding aortic clamping. We present the first reported case of ASGI due to Streptococcus haemolyticus, the second case due to a fungus, and the second reported case of spondylodiscitis after EVAR.

 

Endovascular treatment of aortic abdominal aneurysms (EVAR) has gained in popularity in recent years. EVAR follow-up is primarily focused on technical aspects of the procedure (endoleaks, migration of device, neck dilation, and rupture) that have been widely analyzed and reported.1 Nevertheless, due to the low frequency of aortic stent-graft infections (ASGIs), little clinical, treatment, and surveillance data are available.

Most descriptions of ASGI in the literature are in the form of single case reports. Four recent retrospective studies have shown data on the frequency and have increased our knowledge of ASGI.1, 2, 3, 4 These studies have reported a high overall mortality rate of 20-40% and have claimed that the best treatment is endograft removal. However, successful conservative treatment has been reported in some cases.

In this report, we examine the diagnosis, treatment, and follow-up of two cases of ASGI in our center. Despite their undergoing two different therapeutic approaches, both were successful.

Back to Article Outline

Case Reports 

Since 1998, 167 patients have undergone elective EVAR for aortic abdominal aneurysm (AAA) in our university hospital center. All these patients were followed in an outpatient clinic at 1, 6, and 12 months and thereafter once a year. At the first visit, patients underwent anteroposterior and lateral plain abdominal X-ray and contrast-enhanced computed tomography (CT). In this way, the integrity of the stent and presence of endoleaks were assessed. At the second visit, patients underwent another anteroposterior and lateral plain abdominal X-ray and a duplex scan measuring neck, aneurysm, and iliac diameters. If an endoleak was not observed or the aneurysm diameters were stable or reduced, another X-ray and new duplex imaging were performed consecutively at each visit.

During follow-up, two cases with ASGI were diagnosed (1.2%). These two patients underwent scheduled EVAR in the same month in 2007 and were followed at 1 and 6 months without complications. Both received a single dose of cefazolin 2g as antibiotic prophylaxis in surgery.

Patient 1 

Patient 1 was a 76-year-old man with a past history of diabetes mellitus type II, arterial hypertension, high cholesterol level, chronic obstructive pulmonary disease, and myocardial infarction. He underwent first an uneventful elective embolization of the right internal iliac artery and, 1 month later, an elective EVAR using an aortobi-iliac Excluder stent graft (W.L. Gore, Flagstaff, AZ).

Six months afterward, he presented with back pain, 12kg weight loss in the preceding 3 months, and purulent secretion through the right groin. Physical examination showed a firm and tender mass palpable in the right iliac fossa and a cutaneous groin fistula. Blood samples showed a leukocyte count of 10.5×109/L. CT scanning revealed destructive spondylodiscitis L3–L4 and an abscess formation around the aneurysm, in the bilateral psoas muscle compartment, the pelvis, and the anterior abdominal wall in continuity with the right groin (Fig. 1, Fig. 2). This latter abscess was drained by echoguided puncture. The culture of the aspirate revealed Streptococcus haemolyticus and Propionibacterium. A tagged white blood cell scan was performed and showed no uptake in the endograft. Due to patient morbidity, we decided to attempt a conservative treatment with debridement of the abscess, intravenous antibiotic therapy, and lumbar immobilization. Through a right retroperitoneal approach, careful debridement of the abscess and lavage were performed. We observed a ruptured right external iliac artery, which was completely covered by the endograft. Considering that the abscess was an infected hematoma secondary to traumatic arterial rupture and there was no concentration of radiolabeled leukocytes in the endograft, opening the aneurysm sac was not deemed appropriate. Two Penrose drains were placed in the cavity. Cultures of the secretion revealed S. haemolyticus. Postoperatively, saline irrigation via retroperitoneal drains was performed. No bacteria grew in the drain liquid. The patient received first an intravenous empirical antibiotic, vancomycin 1g every 12hr for 10 days, and later appropriately tailored broad-spectrum intravenous antibiotics, sulfamethoxazole 400mg and trimetoprim 80mg every 8hr, rifampicin 900mg a day, and cefuroxime 500mg every 12hr for 15 days. Improvements in the general condition and white blood cell count were noted. The cutaneous fistula and surgical wound healed by secondary intention. The patient was discharged 11 days later with an oral antibiotic, linezolid 600mg every 12hr for 3 months. A CT scan 6 months later confirmed disappearance of the abscess (Fig. 3, Fig. 4).

  • View full-size image.
  • Fig. 1 

    Right arrow indicates an abscess in the psoas muscle compartment and around the aortic aneurysm. Left arrow indicates abscess formation in the anterior abdominal wall.

Patient 2 

A 77-year-old man with a past history of arterial hypertension, high cholesterol level, and myocardial infarction underwent elective EVAR using a left aortouni-iliac Talent device (Medtronic Vascular, Santa Rosa, CA), occlusion of the right common iliac artery, and left-to-right femorofemoral bypass with 8mm ringed polytetrafluorothylene (PTFE). Eleven months after, he presented with back pain, asthenia, anorexia, 10kg weight loss, and fever in the last 2 days. On physical examination, the abdomen was tender in the epigastric area to deep palpation. His blood results showed an elevated white blood cell count to 16×109/L with 75% of neutrophils and anemia of 7.5g/dL. An abdominal X-ray did not show abnormalities. A CT scan revealed the stent graft surrounded by an abscess with air and in intimate contact with the third duodenal portion (Fig. 5). Tc-99m leukocyte scintigraphy demonstrated marked concentration of radiolabeled leukocytes in the central region of the abdomen, at the level of the L3 vertebra (Fig. 6). Endoscopy showed a hole in the third duodenal portion with blood flowing through it, confirming the diagnosis of aortoenteric fistula. He underwent an emergency operation. First, a left-sided axillofemoral bypass was performed, employing an 8mm PTFE prosthesis. Through a midline laparotomy the endograft was completely removed, and the infrarenal aorta and left common iliac artery were closed using nonresorbable sutures and covered with omentum. Instead of the hole of the third duodenal portion, a fibrous area could be observed, and we decided not to perform intestinal bypass and, thus, to cover the fibrous area with omentum. Finally, debridement of the abscess and lavage of the cavity were performed. Streptococcus intermedius grew in two blood cultures, and Candida albicans grew in the endograft culture. Nevertheless, microbiological cultures of aortic tissue and the abscess were negative. Firstly, the patient received an intravenous empirical antibiotic, vancomycin 1g every 12hr. After obtaining blood culture results, intravenous piperacillin 4g and tazobactam 0.5g every 8hr were administered to the patient during hospitalization. The postoperative course was uneventful, and the patient was discharged in good clinical condition receiving oral amoxicillin/clavulanate 875mg and fluconazole 200mg every 8hr.

Five days after discharge, the patient was admitted to hospital with fever and back pain. Blood samples showed a leukocyte count of 15.5×109/L. Emergency CT scan showed fluid collection with air surrounding the aortic stump and the third duodenal portion (Fig. 7). Endoscopy revealed again a hole in the third duodenal portion. Enterobacter cloacae grew in blood cultures. To avoid the aggressiveness of a second surgery, we decided to treat him with an empirical antibiotic, meropenem 500mg every 8hr for 12 days. The clinical symptoms improved, as did the CT scan appearance. A month later the abscess was reduced to a minimum and air in the CT scan disappeared. The patient was discharged, receiving an appropriately tailored broad-spectrum intravenous antibiotic, moxifloxacin 400mg every 12hr for 1 month.

Six months after surgery, both the clinical syndrome and CT scan showed no evidence of recurrent infection (Fig. 8).

Back to Article Outline

Discussion 

ASGIs are uncommon and frequently associated with high mortality and morbidity rates.5 The most recent retrospective study showed a 0.2% rate of ASGI after 494 EVAR procedures, with an overall mortality rate of 27.3% in the older study and 40% in the most recent one.1, 2

Specific predisposing risk factors could not be identified from these small published series; however, some aspects seem to increase the likelihood of ASGI. The unevacuated thrombus, persistent endoleak, performance of procedures in an interventional radiology suite,4 multiple endovascular or surgical procedures at follow-up,2, 3 nosocomial bloodstream septicemia,6 and immunocompromised states1 could be related to ASGI. Both cases were performed in a conventional operating room, had no immunosuppression, and presented no signs of bloodstream septicemia previously or endoleaks in the previous CT. Nevertheless, patient 1 underwent an adjuvant endovascular procedure. In retrospective studies,2, 4 30.6-35.5% of EVAR procedures were complicated by infection and associated with at least one additional endovascular procedure. Successive percutaneous puncture of the skin could explain that the most common causative agents of ASGI are skin microorganisms.1, 7 In the studies mentioned, Staphylococcus aureus is the infectious agent most commonly implicated in ASGI. In the abscess culture of patient 1 grew Streptococcus haemolyticus and Propionibacterium, two bacteria that usually contaminate the skin. To our knowledge, this is the first reported case of endograft infection due to S. haemolyticus. Some centers routinely cover secondary interventions in EVAR patients by a single dose of teicoplanin and cefuroxime.3 In patient 2 S. intermedius and C. albicans grew firstly in the blood and endograft cultures, respectively. Both bacteria are commensals of the gastrointestinal tract. It is often dificult to determine whether duodenal defects expose endografts to bacteria or whether a previous infection of the endograft caused erosion of the third duodenal portion wall. Sharif et al.3 suggested the latter in their report. To our knowledge, this is the second case of infected endograft due to a fungus. The first one died from septic shock after surgery.8

Since both patients had been operated in the same month, the sterilization protocols of that particular month were specifically reviewed, without finding any flaw.

There is no consensus regarding which is the most resistant to infection of the prosthesis. Whereas one experimental in vitro study reported that expanded PTFE (ePTFE) grafts might be more resistant to infection than Dacron,9 another in vivo study reported similar ratios of colonization and infection of ePTFE and Dacron.9, 10 The prosthesis of patient 1 was made of ePTFE. Patient 2 received a Dacron stent graft that was infected by fungus. Some experimental studies on canine models have reported that the stent graft is more susceptible to infection in the early postoperative period and becomes less susceptible 1 week after implantation,11 hence the importance of antibiotic prophylaxis. Some in vitro reports have suggested that antimicrobially coated vascular grafts may protect against infection by organisms which can perioperatively colonize the vascular graft but may not cause clinical findings of infection until a long time after.10

Patient 1 is the second reported case of spondylodiscitis after EVAR. The first reported case was treated by removal of the endograft and necrotic fragments of the former L3-L4 disk and recovered well. In our patient, due to his high surgical risk, we decided to treat spondylodiscitis conservatively. Some reports have supported that prolonged antibiotic therapy is essential for cure in most patients, and has contributed to reduce morbidity and mortality.12

The diagnosis of ASGI could be a challenging problem. Patients usually present with systemic unspecific clinical syndrome (malaise, fever, weight loss, back or abdominal pain). These symptoms, with CT findings such as fluid collection around the aorta or indwelling air within the aneurysmal sac, could have made the diagnosis, as in our two patients. In patient 2, a tagged white blood cell scan could be perfomed to support the diagnosis. Diagnosis of spondylodiscitis in CT findings could be difficult as noninfectious vertebral destruction has been described after EVAR and pseudoaneurysm.

Unfortunately, there are no guidelines for the best treatment of ASGI. In most reports, complete infected stent graft removal with or without extra-anatomical bypass and intravenous antibiotics is clearly advised. Fiorani et al.2 reported an operative mortality rate of 16.3%. Ducasse et al.4 reported a mortality rate of 16% after employing extra-anatomical bypass and a 5.8% rate after in situ reconstruction. Conservative treatment with suppressive antibiotic therapy and abscess drainage has been reported; however, it is not largely advocated as the best therapeutic approach.4, 13 Fiorani et al.2 reported a conservative treatment mortality rate of 36.4%, Ducasse et al.4 reported 36.4%, and Heyer et al.1 reported 43%. Even so, in patients with high surgical risk, such as patient 1, conservative treatment could be justified to avoid suprarenal aortic clamping, as shown recently. Both were patients with multiple cardiorespiratory comorbidities. Radiological drainage was chosen in one case and open drainage in the other case.14, 15

Consensus regarding the duration of antibiotic treatment has yet to be reached. In our patients the duration of antibiotic treatment was based on multidisciplinary consensus involving infectious disease specialists and vascular surgeons.

Back to Article Outline

Conclusion 

ASGIs are uncommon but associated with high mortality. The two forms of treatment available, conservative treatment versus complete infected stent-graft removal, should be tailored according to the patient comorbidities.

Back to Article Outline

References 

  1. Heyer K, Modi P, Morasch M, et al. Secondary infections of thoracic and abdominal aortic endografts. J Vasc Interv Radiol. 2009;20:173–179
  2. Fiorani P, Speziale F, Calisti A, et al. Endovascular graft infection: preliminary results of an international enquiry. J Endovasc Ther. 2003;10:919–927
  3. Sharif M, Bernard L, Lau L, et al. Prosthetic stent graft infection after endovascular abdominal aortic aneurysm repair. J Vasc Surg. 2007;46:442–448
  4. Ducasse E, Calisti A, Speziale F, et al. Aortoiliac stent graft infection: current problems and management. Ann Vasc Surg. 2004;18:521–526
  5. Kitamura T, Morota T, Motomura N, et al. Management of infected grafts and aneurysms of the aorta. Ann Vasc Surg. 2005;19:335–342
  6. Vogel TR, Symons R, Flum DR. The incidence and factors associated with graft infection after aortic aneurysm repair. J Vasc Surg. 2008;47:264–269
  7. Koning H, Van Sterkenburg S, Pierie M, et al. Endovascular abdominal aortic aneurysm repair complicated by spondylodiscitis and iliaco-enteral fistula. J Vasc Surg. 2008;47:1330–1332
  8. Bakoyiannis CN, Georgopoulos SE, Tsekouras NS, et al. Fungal infection of aortoiliac endograft: a case report and review of the literature. Ann Vasc Surg. 2007;21:228–231
  9. Zdanowski Z, Ribbe E, Schalén C. Bacterial adherence to synthetic vascular prostheses and influence of human plasma. An in vitro study. Eur J Vasc Surg. 1993;7:277–282
  10. Darouiche RO, Mansouri MD. In vitro activity and in vivo efficacy of antimicrobial-coated vascular grafts. Ann Vasc Surg. 2004;18:497–501
  11. Kirksey L, Brener BJ, Hertz S, et al. Prophylactic antibiotics prior to bacteremia decrease endovascular graft infection in dogs. Vasc Endovasc Surg. 2002;36:171–178
  12. Pintado-García V. Infectious spondylitis. Enferm Infecc Microbiol Clin. 2008;26:510–517
  13. Zegelman M, Gunther G. Infected grafts require excision and extra-anatomic reconstruction: against the motion. In:  Greenhalgh R editors. The Evidence for Vascular or Endovascular Reconstruction. Philadelphia: WB Saunders; 2002;p. 235–245
  14. Hulin SJ, Morris GE. Aortic endograft infection: open surgical management with endograft preservation. Eur J Vasc Endovasc Surg. 2007;34:191–193
  15. Saleem B, Berger P, Zeebregts C, et al. Periaortic endograft infection due to Listeria monocytogenes treated with graft preservation. J Vasc Surg. 2008;47:635–637

PII: S0890-5096(09)00343-4

doi:10.1016/j.avsg.2009.11.003

Annals of Vascular Surgery
Volume 24, Issue 4 , Pages 554.e1-554.e5, May 2010

Article Tools

* Image Usage & Resolution