Background: Surgery and high-dose radioactive iodine ((131)I) treatment are the cornerstones in the treatment of differentiated thyroid cancer. Patients without (131)I uptake on the post-therapeutic whole body scan (WBS), but with detectable thyroglobulin (Tg) during thyroxine withdrawal (Tg-off), are evaluated with an 18-fluorodeoxyglucose positron emission tomography ((18)F-FDG-PET) for tumor localization within three months. The yield of (18)F-FDG-PET imaging and clinical usefulness of a Tg-off cutoff value to predict a positive scan were assessed.
Methods: From 2002 to 2011, 52 patients with a negative WBS and concurrent detectable Tg-off were evaluated. Thirty-five PET scans were performed during initial treatment, 17 after recurrent disease. Thirty-two patients were on substitution therapy, 17 were evaluated with endogenous thyrotropin elevation, and 3 after recombinant human thyrotropin stimulation. To determine the Tg-off cutoff value, a receiver operating characteristic curve was used.
Results: Nine (17%) (18)F-FDG-PET scans were true positive, 3 (6%) false positive, 36 (69%) true negative, and 4 (8%) false negative (sensitivity 69%, specificity 92%). In 13%, a true-positive scan resulted in a change in the clinical management. The area under the receiver operating characteristic curve is 0.82 [CI 0.64-0.99] (p<0.01), and the Tg-off cutoff value is 38.00 ng/mL (sensitivity 67%, specificity 95%). Ninety percent of (18)F-FDG-PET true-positive patients had a Tg-off >2.00 ng/mL.
Conclusions: An (18)F-FDG-PET within three months after a negative WBS with detectable Tg-off showed additional tumor localization in 17% of the patients, leading to a change in clinical management in 13%. A clinically useful Tg-off cutoff value was not found, but 90% of positive (18)F-FDG-PET scans occurred in patients with a Tg-off >2.00 ng/mL.