Objectives We aimed to clarify the role of FDG-PET in the management of large-vessel vasculitis, focusing on three issues i) describe and determine the different FDG-PET criteria for the diagnosis of vascular inflammation ii) the performance of FDG-PET for the diagnosis of large-vessel inflammation in GCA patients and iii) the performance of FDG-PET to evaluate the disease inflammatory activity in TA patients.
Methods MEDLINE, Cochrane Library and EMBASE database were searched for articles that evaluated the value of FDG-PET in large-vessel vasculitis, from 01/2000 to 12/2013. Inclusion criteria were: American College of Rheumatology (ACR) criteria for GCA or TA, definition PET positivity threshold, and >4 cases included. Sensitivity (Se) and specificity (Sp) of FDG-PET for the diagnosis of large-vessel inflammation were calculated from each included individual study, and then pooled for meta-analysis with a random effects model.
Results 21 studies (413 patients, 299 controls) were included in the systematic review. FDG-PET showed FDG vascular uptake in 70% (288/413) of patients and 7% (22/299) of controls. Only vascular uptake equal or higher than the liver uptake was significantly different between GCA/TA patients and controls (p<0.001). The meta-analysis of GCA patients (4 studies, 57 patients), shows that FDG-PET has high Se and Sp for the diagnosis of large-vessel inflammation in GCA patients in comparison to controls, with a pooled Se at 90% (95%CI,0.79-0.96) and a pooled Sp at 98% (95% CI,0.94-0.99). The meta-analysis of TA patients (7 studies, 191 patients), shows that FDG-PET has a pooled Se at 87% (95%CI,0.78- 0.93) and Sp at 73% (95%CI,0.63- 0.81) for the assessment of disease activity in TA, with Sp to 84%, with studies using NIH criteria as disease activity assessment scale.
Conclusions FDG-PET showed good performances for the diagnosis of large-vessel inflammation, with higher accuracy in GCA compared to TA patients. A visual analysis of vascular uptake, equal or higher than the liver uptake, appears as the best criteria for the diagnosis of large-vessel vasculitis.
Disclosure of Interest None declared