Large-Vessel Vasculitis: Interobserver Agreement and Diagnostic Accuracy of (18)F-FDG-PET/CT

Introduction. (18)F-FDG-PET visualises inflammation. Both atherosclerosis and giant cell arteritis cause vascular inflammation, but distinguishing the two may be difficult. The goal of this study was to assess interobserver agreement and diagnostic accuracy of (18)F-FDG-PET for the detection of large artery involvement in giant cell arteritis (GCA). Methods. 31 (18)F-FDG-PET/CT scans were selected from 2 databases. Four observers assessed vascular wall (18)F-FDG uptake, initially without and subsequently with predefined observer criteria (i.e., vascular wall (18)F-FDG uptake compared to liver or femoral artery (18)F-FDG uptake). External validation was performed by two additional observers. Sensitivity and specificity of (18)F-FDG-PET were determined by comparing scan results to a consensus diagnosis. Results. The highest interobserver agreement (kappa: 0.96 in initial study and 0.79 in external validation) was observed when vascular wall (18)F-FDG uptake higher than liver uptake was used as a diagnostic criterion, although agreement was also good without predefined criteria (kappa: 0.68 and 0.85). Sensitivity and specificity were comparable for these methods. The criterion of vascular wall (18)F-FDG uptake equal to liver (18)F-FDG uptake had low specificity. Conclusion. Standardization of image assessment for vascular wall (18)F-FDG uptake promotes observer agreement, enables comparative studies, and does not appear to result in loss of diagnostic accuracy compared to nonstandardized assessment.