Diagnostic performance of different quantitative and qualitative parameters and their combined value for the diagnosis of aortic graft infection using [18F]-FDG PET/CT

224 Background: The aim of this study was the evaluation of different quantitative and qualitative parameters and their combined value for the diagnosis of aortic graft infection (AGI) applying [18F]fluorodeoxyglucose (FDG) positron emission tomography (PET)/computed tomography (CT). Methods: [18F]-FDG PET/CT was performed in 50 patients with clinically suspected AGI. In addition, 12 oncological patients with aortic repair but without AGI were included in the analysis. The [18F]-FDG uptake pattern around the graft was assessed using a five-point visual grading score (VGS) next to maximum standardized uptake values (SUVmax) and graft-to-background ratios (GBRs) with blood pool and non-inflammatory aortic wall as a reference. The diagnostic performance of VGS, SUVmax and different GBRs was assessed by receiver-operating-characteristic analysis (ROC). The ‘Management of Aortic Graft Infection Collaboration’ criteria served as the standard of reference for the final diagnosis. Results: 28 infected and 34 uninfected grafts were identified. SUVmax was the most powerful predictor for the diagnosis of AGI according to the ROC analysis (AUC of 0.988; cut off value of >4.96), resulting in a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 89.3%, 100%, 100%, 91.9% and 95.2%, respectively. Notably, there was no benefit using GBRs compared to SUVmax findings. The VGS (AUC of 0.983; cut off value of ≥3) resulted in a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 100%, 85.3%, 84.8%, 100% and 91.9% respectively. Thus, both quantitative and qualitative approaches provide complementary diagnostic information with congruent and accurate results in overall 54/62 (87.1%) patients. Additionally, PET/CT provided a conclusive diagnosis in 5 patients with initially suspected but finally ruled out AGI (i.e. other sites of infection/inflammation). Conclusions: Quantitative analysis by SUVmax in combination with qualitative analysis by VGS provides a good reference to assess AGI and may increase the diagnostic accuracy, confidence and acceptance of [18F]-FDG PET/CT in the setting of suspected graft infection.