Objective: To evaluate the predictive value of clinical and biochemical features when compared to 18FDG-PET in the diagnostic work-up of large vessel vasculitis (LVV).
Methods: Twenty-five patients underwent 18FDG-PET for the clinical suspect of LVV. All of them presented history of systemic symptoms lasting >or=6 months and laboratoristic evidence of persistently high markers of inflammation. The patients were stratified according with: i) clinical manifestations, defined as presence of one or more ACR criteria for the classification of LVV; ii) laboratory investigations: Erythrocyte Sedimentation Rate (ESR) higher or lower than 50 mm/h, C-Reactive Protein (CRP) higher or lower than 2 mg/dl; iii) prednisone dose in the 4 weeks preceding PET examination.
Results: The total number of positive PET was higher in the group without clinical ACR criteria and in the group with inflammation markers under the established cut-off. The number of scans consistent with LVV was higher in the groups presenting one or more clinical criteria for LVV but in those with very high ESR and CRP. In all the cases differences between groups were not statistically significative. A clear cut negative correlation between steroid dose and number of scans suggestive for LVV has been observed.
Conclusions: Diagnosis of LVV remains challenging, especially in patients presenting with a constellation of non-specific symptoms and laboratory findings. In this study, both clinical and biochemical features show low correlation with a vasculitic pattern of FDG uptake. In our experience 18FDG-PET represents an useful diagnostic tool in early stages of LVV and a powerful instrument to follow the treatment responses.