Introduction
Giant cell arteritis (GCA) is a frequent granulomatous vasculitis. In the diagnosis of large vessel GCA, FDG-PET/CT imaging has been recommended and its use has steadily increased. The objective of this prospective study was to evaluate the diagnostic performance of digital FDG-PET imaging with a head and neck dedicated protocol, especially to detect cranial arteries involvement, using different visual grading scores, and to assess the inter-reader agreement.
Material and Methods
FDG-PET/CT scans with a 10-min acquisition time on the head and neck were performed in patients suspected of having recent GCA. A score including 20 segments due to the inclusion of cranial segments and a score established only for the cranial segments were compared to the Total Vascular Score (TVS) established on 7 arterial segments. These scores were assessed by two nuclear medicine physicians on a blinded manner and, in case of disagreement, a consensus was established with a third nuclear medicine physician.
Results
GCA diagnosis was retained for 15 of the 52 patients included. Sensitivity and specificity of the 20-segment or the cranial scores were higher than those of TVS (sensitivity and specificity of 80% of 97% for both scores vs 73% and 89% for TVS, respectively). The inter-reader agreement was substantial in the 20-segment score (weighted kappa = 0.68) but better in the cranial segments than other segments.
Conclusion
This study shows the usefulness of grading scores able to highlight involvement of cranial arteries in GCA with a better inter-reader agreement in cranial segments than other segments.