Diagnostic accuracy of ultrasound for detecting large-vessel giant cell arteritis using FDG PET/CT as the reference.

Abstract

The diagnostic accuracy of axillary artery US in the diagnosis of large-vessel (LV)-GCA using 18F-fluorodeoxyglucose (FDG) PET/CT as reference standard was prospectively evaluated in GCA-suspected patients. As an exploratory analysis, the diagnostic accuracy of cranial artery FDG PET/CT was evaluated. Briefly, the inclusion criteria were age ≥50 years, raised inflammatory markers and potential GCA symptoms. Patients in immunosuppressive therapy or with a previous diagnosis of GCA or PMR were excluded. Examinations were performed pre-treatment. LV-GCA reference diagnosis was a clinical diagnosis of GCA and PET-proven LV inflammation. GCA patients fulfilling ACR criteria were considered as cranial-GCA (c-GCA). Patients without GCA were considered controls. Receiver operating characteristic curve analysis of the US-measured axillary intima-media thickness was performed. FDG uptake in temporal, maxillary and vertebral arteries was also assessed. Forty-six patients were diagnosed with LV-GCA, 10 with isolated c-GCA, and in 34 patients GCA was dismissed. Axillary US yielded a sensitivity of 76% and a specificity of 100% for LV-GCA. An axillary intima-media thickness cut-off of 1.0 mm yielded a sensitivity of 74% and a specificity of 92%. Adding LV US to temporal assessment increased sensitivity from 71% to 97% (all GCA patients). Cranial artery PET showed a diagnostic sensitivity of 78% and specificity of 100% for c-GCA. Axillary artery US shows high accuracy for the LV-GCA diagnosis. Building upon the recent EULAR recommendations, we propose a diagnostic algorithm with US as the first-line confirmatory test, not only in c-GCA-suspected patients, but in all patients suspected of GCA.