The increasing use of endovascular aneurysm repair (EVAR) appears to be associated
with the burden of vascular endograft infections. Complete stent graft explantation
is recommended but leads to significant mortality. This study aimed to assess the
technical challenges, complications, and mortality rate following infected endograft
Patients who underwent abdominal aortic endograft explantation for infection at the
Bordeaux University Hospital from July 2008 to December 2020 were included retrospectively
in this single centre observational study. The diagnosis was established based on
the MAGIC criteria. The primary endpoint was 30 day mortality. Secondary endpoints
were 90 day and in hospital mortality, survival, and re-infection.
Thirty-four patients were included, median age 69 years (interquartile range [IQR]
65, 76), with four (12%) treated as an emergency. The median time from EVAR to explantation
was 17.5 months (4.5 – 36.3). In situ reconstruction was carried out with prosthetic grafts in 24 patients (71%, including
23 antimicrobial grafts combining silver and triclosan), and biological grafts in
10 (five femoral veins, four arterial allografts, three bovine patches, one biosynthetic
graft). Seventeen aorto-enteric fistulae (AEnF) were addressed with direct repair
of the intestinal tract (n = 10/17; 59%) or resection and anastomosis (n = 7/17; 41%). The culture was polymicrobial in 12 patients (35%) and remained sterile
in four (12%). The 30 day and in hospital mortality rates were 21% (n = 7) and 27% (n = 9). Twenty-five patients (73%) presented with early post-operative complications,
requiring 16 revision procedures (47%). Over a median follow up of 16.2 months (IQR
8.3, 33.6), the mortality rate was 35% (n = 12; 11 aortic related; 32%), with two re-infections (6%), both after biological
reconstruction (one for an AEnF).
Early morbidity and mortality remain high after complete infected endograft explantation,
even in a high volume centre. Comparison with other treatment modalities in large
multicentre cohorts might be of interest.