Graft infections caused by fungus are quite rare, and the incidence reported in the literature is less than 10% of all graft infections [5, 8, 9]. The number of reported cases of graft infections by Aspergillus species was only 15 or less between 1980 and 2001 [10, 11].
Detection of pathogenic microorganisms is a gold standard for correct diagnoses and appropriate therapies for graft infections. However, specifically, Aspergillus species are known to be poorly detectable in blood culture studies [10,11,12,13]. Instead, a fungal antigen marker βDG can detect invasive Aspergillosis with sensitivity of 76%. When combined with Aspergillus GM antigen test, specificity of the test increases to 98% [14], which can be useful in making clinical decisions. In the present case, we used these tests to determine the pathogenic microbes. The positive Aspergillus GM antigen test as well as elevation of βDG values along with clinical symptoms provided evidence for Aspergillus infection, which enabled us to select anti-fungal drug with appropriate spectrum even though the multiple culture studies were negative.
To detect the infectious loci, we first narrowed the target to the descending aorta depending on the distribution of repetitive embolisms. The patient experienced recurrent embolisms in the areas fed by the SMA and splenic artery but never in cerebral arteries. Therefore, we excluded the involvement of the ascending aortic graft as an origin of emboli. Infectious endocarditis (IE) was also excluded by the same reason though this patient had a history of MVR. The CT and [18F]FDG PET/CT played important roles in detecting the infectious loci. [18F]FDG PET/CT is known to be useful to detect foci of graft infections with a sensitivity of 85% and a specificity of 68.4% [15]. Unlike our patient, most cases with graft infection had their infectious foci in the ascending aortic or arch grafts, which was probably because of more turbulent flow in these areas. We considered the atypical infectious foci in our case were because of the unique structural features of the patient’s aorta. In particular, the descending aortic graft was connected to the patient’s own aorta, forming a nested structure, and their junction was bent for the graft’s elasticity, as observed in the CT image (Fig. 2A, right). A turbulent blood flow caused by these complex structures may be an explanation for fungal growth at the site.
This case is also notable for remote onset of infection after an aortic graft insertion. In many reports, end-graft infections occurred during the early phase after insertion of the graft [16]. However, a year before the first admission, the patient had received MVR. Thus, there is a possibility that the open chest surgery caused the graft infection. MVR has been reported to cause IE, and cumulative IE risk after MVR within 10 years is reported to be 5.2% [17].
Current therapeutic options for graft infections are surgical and medication therapies [18]. Surgical interventions may be more curable than medication therapies, but they are excessively invasive. Most vascular grafts were not designed to be exchanged after the first placement, and one study showed that the mortality rate was 18–30% when grafts were surgically replaced [18]. Regarding graft infections after TEVAR/EVAR, another study in 2011 reported there was no significant difference in mortality rate between surgical and medication therapies [19]. On the other hand, if surgical replacement is not applicable, patients need to receive long-term broad-spectrum antibiotic therapy, which could cause drug-induced complications [20]. For those reasons, interventional strategies should be critically determined depending on conditions and underlying diseases of each patient. In our case, we thought a surgical approach was the most appropriate way because: (1) the patient was still tolerant for surgical replacement of the infected graft; and (2) oral anti-fungal medication could not prevent the recurrence of arterial embolism.
In conclusion, we experienced a rare case of repeated large-arterial embolisms originating from late-onset end-graft infection in the descending aorta. For the management of complex graft infection as in the present case, this report highlights (1) the importance of repeated imaging studies by using all available modalities including [18F]FDG PET/CT to specify the infectious loci, and (2) the usefulness of non-cultural tests for diagnosis and selection of effective agents, especially in case of fungal infection.