Diagnosis of Aortic Graft Infection

Presentation

The manifestation of an aortic graft infection depends on the interval between graft implantation and the onset of the infection, the nature of the organism, and the anatomic location of the prosthesis. Early-appearing graft infections, occurring less than 4 months after implantation, are more likely to involve the graft diffusely and to have systemic manifestations. This occurs either because the prosthesis is not yet isolated by the perigraft capsule or perhaps because of the nature of the most commonly involved organisms: Staphylococcus aureus and gram-negative organisms. The enzymes associated with S. aureus infection result in a suppurative inflammatory reaction, facilitating local tissue invasion. Gram-negative organisms are somewhat less virulent, but they also cause suppuration and often abscess formation. Pseudomonas species, in particular, demonstrate vascular wall invasion, pseudoaneurysm formation, and occasionally arterial wall disruption. In contrast to these organisms, Staphylococcus epidermidis, a very common causative organism in late-appearing graft infections, is a low-virulence organism that produces an indolent infection that usually manifests years after prosthesis implantation. Although the involvement of the graft by these lower-virulence organisms may be either focal or diffuse, patients are more likely to manifest only focal findings of localized fluid collections (with or without associated cellulitis), recurrent draining wounds, or chronic sinus tracts.

In general, aortic graft infections tend to manifest late and have a much broader potential spectrum of symptoms ( Table 1 ). Infections caused by low-virulence organisms most commonly become manifest by a focal wound infection, a fluid collection, or a draining sinus tract. Systemic symptoms are very uncommon. Infections caused by high-virulence organisms manifest with systemic symptoms, and these can vary from being quite subtle with malaise, discomfort in the region of the prosthesis, or weight loss to the very obvious with fever, leukocytosis, bacteremia, and sepsis. False aneurysms can occur if infection has disrupted an anastomosis.

Aortic graft infections associated with prosthetic–enteric fistulas or erosions should be considered differently because they more likely represent a mechanical problem first, followed by contamination and/or infection of the prosthesis. They manifest late not because of a low-virulence causative organism but because the erosion into the gastrointestinal tract requires time to develop. At least 65% to 75% of erosions and fistulas have some associated gastrointestinal blood loss, but some fistulas and erosions are entirely occult and are discovered only when treatment of the associated graft infection is undertaken.

Prosthetic–enteric fistulas and erosions are most likely to manifest with fever, leukocytosis, bacteremia, and sepsis, but the overall incidence is still low. The pattern of bleeding results from the type of communication with the intestinal tract. Erosions tend to manifest with episodic subacute or chronic low-grade bleeding because the source of bleeding is the eroded edge of the bowel wall. Fistulas tend to manifest with acute bleeding, including the typical herald bleeding episode, because the source of bleeding is the arterial lumen itself. When the gastrointestinal tract communication is associated with an anastomotic false aneurysm, infection may be the primary cause of the false aneurysm, with the intestinal tract erosion secondary to the mass effect of the false aneurysm. Occult graft infection manifesting as acute ischemia secondary to graft or graft limb occlusion is rare, as is septic embolization.

Diagnosis

The diagnosis of aortic graft infection is the first of the many challenges posed by this clinical condition. The goal of the diagnostic evaluation is to establish both the presence of the infection and the extent of graft involvement. It is important to remember that all modalities used to evaluate aortic graft infection are inferential, because none can actually see infection. Rather, each modality looks for changes that are the consequences of infection. The ideal diagnostic method would detect a change that is unique to the presence of infection. No such diagnostic modality exists, and therefore it is common to use multiple modalities, with the goal of increasing the overall diagnostic accuracy. A positive test is clinically useful; a negative test is less useful. A negative test is less useful, because the false-negative rates for all diagnostic modalities remain high.

Diagnostic modalities can be categorized as either anatomic or functional. The anatomic modalities are computed tomography (CT), magnetic resonance imaging (MRI), sinography, and endoscopy. The functional modalities include labeled white cell scans (using indium or technetium-99m), and fluorodeoxyglucose–positron emission tomography (FDG-PET) ( Table 2 ) .