Background/Objectives: Imaging studies have transformed the diagnosis of large vessel vasculitis (LVV) involvement in giant cell arteritis (GCA). A positron emission tomography/computed tomography (PET/CT) scan with 18-fluorodeoxyglucose (18F-FDG) has emerged as a valuable tool for assessing LVV. We aimed to determine the utility of an 18F-FDG-PET/CT scan in detecting LVV in GCA in the ARTESER registry. Methods: The ARTESER study is a large multicenter, retrospective, longitudinal, and observational study, promoted by the Spanish Society of Rheumatology. It included patients newly diagnosed with GCA across 26 tertiary hospitals from 1 June 2013 to 29 March 2019. Patients with a diagnosis of incidental GCA were included if they fulfilled specific criteria, including the ACR 1990 criteria, positive imaging examinations, or the expert clinical opinion of investigators. Differences between patients with positive and negative 18F-FDG-PET/CT scan results were analyzed using a bivariate model. A regression model assessed associations in patients with a positive scan, and the predictive capacity of the cumulative dose of glucocorticoids (GC) on PET scan outcomes was evaluated using ROC curve analysis. Results: Out of 1675 GCA patients included in the registry, 377 met the inclusion criteria of having an 18F-FDG-PET/CT scan. The majority were diagnosed with a cranial GCA phenotype, and 65% had LVV. The thoracic aorta was the most frequently affected. Cardiovascular disease, diabetes, and older age had a negative association with a positive scan outcome. The OR for having a positive 18F-FDG-PET/CTC scan was lower as the number of days increased. Depending on the cumulative dosage of the GC, the 18F-FDG-PET/CT scan showed an AUC of 0.74, with a Youden index > 60 mg/day. Conclusions: Younger patients showed a higher probability of presenting LVV as detected by the 18F-FDG-PET/CT scan. The timing of the examination and the cumulative dosage of the GC influenced the likelihood of a positive result, with earlier tests being more likely to detect inflammation.
Read full abstract