Simple dichotomous assessment of cranial artery inflammation by conventional 18F-FDG PET/CT shows high accuracy for the diagnosis of giant cell arteritis: a case-control study

Publikation: Bidrag til tidsskrift/Konferencebidrag i tidsskrift /Bidrag til avisTidsskriftartikelForskningpeer review


PURPOSE: To estimate the diagnostic accuracy of conventional 18F-FDG PET/CT of cranial arteries in the diagnosis of giant cell arteritis (GCA).

METHODS: The study was a retrospective case-control study. The reference diagnosis was fulfillment of the 1990 ACR criteria for GCA. All patients had new-onset GCA. Conventional 18F-FDG PET/CT was performed before glucocorticoid treatment. Controls were age- and sex-matched patients with a previous history of malignant melanoma (MM) undergoing surveillance PET/CT >6 months after MM resection. PET images were evenly cropped to include only head and neck and were assessed in random order by four nuclear medicine physicians blinded to reference diagnosis. Temporal (TA), maxillary (MA) and vertebral (VA) arteries were visually rated for 18F-FDG uptake. Interreader agreement was evaluated by Fleiss kappa.

RESULTS: A total of 44 patients and 44 controls were identified. In both groups, the mean age was 69 years (p = 0.45) and 25/44 were women. 35/41 GCA patients were temporal artery biopsy positive (TAB). Considering only FDG uptake in TA and/or MA, diagnostic sensitivity and specificity was 64 and 100%. Including VA, sensitivity increased to 82% and specificity remained 100%. Interreader agreement was 91% and Fleiss kappa 0.82 for the PET diagnosis based on the cranial arteries.

CONCLUSION: Conventional 18F-FDG PET/CT is an accurate and reliable tool to diagnose cranial arteritis in glucocorticoid-naïve GCA patients. The high diagnostic specificity suggests that TAB can be omitted in patients with 18F-FDG uptake in cranial arteries. 18F-FDG PET/CT performed in patients with suspected vasculitis should always include the head and neck.

TidsskriftEuropean Journal of Nuclear Medicine and Molecular Imaging
Sider (fra-til)184-193
Antal sider10
StatusUdgivet - jan. 2019

Se relationer på Aarhus Universitet Citationsformater

ID: 130630322