The clinical significance of incidental thyroid abnormalities discovered in 18F-FDG PET/CT (FDG PET/CT) studies remains controversial. The objective of this large retrospective study was to a) determine the prevalence of focal FDG thyroid uptake on whole body FDG PET/CT studies performed for non-thyroid cancers and b) to test whether intense focal FDG thyroid uptake is associated with malignancy.
A total of 11,921 FDG PET/CT studies in 6,216 patients performed at our institution between 01/2012 and 12/2014 were analyzed. We retrospectively reviewed the medical records of these patients. Eight hundred forty five/6,216 patients (13.6%) had a thyroid incidentaloma based on the clinical 18FDG PET/CT report. One hundred sixty/845 (18.9%) of these underwent ultrasound and 98 of these (61.3%) had a fine needle aspiration (FNA). Twenty-six of these 98 (26.5%) patients underwent thyroidectomy. Thyroid lesion and background SUV's for each patient were measured upon review of the 18FDG PET/CT study. We measured SUVmax, thyroid to background ( ), thyroid to blood pool ( ) and thyroid to liver ( ) ratios in benign and malignant lesions. Receiver operating curves (ROC) were calculated to determine optimal cut off values between malignant and benign lesions
Twenty-one of the 98 patients who underwent FNA biopsy or thyroidectomy had malignant disease (21.4%). Malignant lesions had significantly higher thyroid lesion SUVmax, , , and than benign nodules. The ROC derived cut off ratio of >2.0 differentiated benign from malignant lesions best with a specificity and sensitivity of 0.76 and 0.88, respectively.