Determining the Diagnostic Value of 18F-Fluorodeoxyglucose Positron Emission/Computed Tomography in Detecting Prosthetic Aortic Graft Infection.

BACKGROUND: To determine the diagnostic value of 18F-fluorodeoxyglucose positron emission/computed tomography (FDG PET/CT) in detecting prosthetic aortic graft infection (AGI).

METHODS: Twenty-one patients with prosthetic grafts for abdominal aortic aneurysms underwent FDG PET/CT scans for suspected graft infection over a 15-month period. Images were evaluated for tracer pattern and grade of FDG uptake in addition to measuring the maximal standardized uptake value (SUVmax). Two independent nuclear medicine physicians retrospectively evaluated all imaging. The images from a control group of patients with aortic grafts who underwent FDG PET/CT scans for onco-hematological indications were evaluated to establish radiological characteristics of asymptomatic grafts. Secondary parameters that are associated with graft infection such as components of the peripheral blood count were collected. Graft infection status was determined using microbiological outcomes following graft explantation or radiological drainage of perigraft collections and correlated with results of the FDG PET/CT scans to determine infective status.

RESULTS: In the control group, the pattern of FDG uptake was homogenous and diffuse. The mean SUVmax was 3.5 (±1.3). Thirteen out of 21 grafts were confirmed as infected. Tracer uptake in infected grafts displayed an intense and focal pattern, with a median grade of uptake of 4 vs. 2 on a validated 4 point grading scale. The area under the receiver operating curve for FDG PET/CT in detecting infection was 0.85 (±0.15) P = 0.01. Sensitivity was 92%, specificity 63%, and positive and negative predictive values of 80% and 83%, respectively. The SUVmax was significantly higher in infected than noninfected grafts, (10.3 ± 4.2 vs. 5.4 ± 3.4) P = 0.02. According to the receiver operating characteristic analysis, SUVmax greater than 6.3 represented the optimal cutoff between infective and noninfective outcome. Of the secondary parameters collected, grade of uptake and SUVmax were the only significant predictors of infection (odds ratio 2.5, 1.5 respectively) P = 0.05. White cell count, erythrocyte sedimentation rate, and C-reactive protein demonstrated nonsignificant odds ratios of 1.4, 0.9, and 1.0, respectively.

CONCLUSIONS: FDG PET/CT is a valuable diagnostic test for identifying AGI. Infected grafts display significantly greater FDG uptake in a distinctive intense focal perigraft pattern and distribution. SUVmax greater than 6.3 is a good cutoff to determine infective status.