18F-fluorodeoxyglucose PET/CT in the diagnosis of vascular graft infection

Abstract

To investigate diagnostic role of 18F-fluorodeoxyglucose PET/CT in patients with suspected vascular graft (VG) infection. A prospective analysis included data of 30 PET/CT examinations for suspected infection of aortic VG (n=27) and bypass grafts (n=3) after surgical treatment (median 48 months). In 77% (23/30) of cases, the diagnosis was initially «possible» (n=11) or «rejected» (n=12) in accordance with common diagnostic criteria. All PET/CT results were verified by clinical, laboratory and intraoperative («=20) data. VG infection was confirmed in 18 patients and ruled out in 12 cases. PET/CT confirmed VG infection in 94% (17/18) and excluded in 50% (6/12) of cases. False PET/CT results were obtained in 23% (7/30) cases: false positive in 6 cases and false negative in 1 case. Thus, sensitivity, specificity and diagnostic accuracy of PET/CT were 94%, 50% and 77%, respectively; positive and negative predictive value - 74% and 86%. PET/CT results allowed correct reclassifying 33% (10/30) of cases. VG infection was confirmed in 73% (8/11) of patients with initially «possible» diagnosis and excluded in 17% (2/12) of patients with initially «rejected» infection. Moreover, whole body PET/CT revealed unknown inflammation foci outside VG in 73% (22/30) of cases. These data were applied to correct treatment approach in 80% (24/30) of cases. Our results showed high efficacy of 18F-fluorodeoxyglucose PET/CT in the diagnosis of VG infection. Despite low specificity, this technique has high sensitivity and accuracy that allowed reclassifying 33% of cases.