Diagnostic utility of PET/CT with 18F-DOPA and 18F-FDG in persistent or recurrent medullary thyroid carcinoma: the importance of calcitonin and carcinoembryonic antigen cutoff

Purpose This study sought to evaluate and compare the utility of 18-F-fluorodihydroxyphenylalanine (F-18-DOPA) and F-18-fluorodeoxyglucose (F-18-FDG) positron emission tomography/computed tomography (PET/CT) for identification of lesions in patients with recurrent medullary thyroid carcinoma (MTC). In addition, we analyzed the correlation between the calcitonin (Ct), carcinoembryonic antigen (CEA) levels, each doubling time (DT), and PET positivity. We evaluated the reliability of the 150 pg/mL Ct cutoff set by the American Thyroid Association guidelines for further imaging (including F-18-DOPA PET/CT). Methods We prospectively recruited 18 patients with recurrent MTC, identified by elevation of Ct or CEA. Each patient underwent a F-18-FDG PET/CT and a F-18-DOPA PET/CT. Results Abnormal uptakes were detected with F-18-DOPA (n=12) and F-18-FDG (n=9), (sensitivity of 66.7% vs. 50%; p<0.01). Twenty-eight lesions were detected with F-18-DOPA vs. 16 lesions with F-18-FDG (1.56 +/- 1.5 vs. 0.89 +/- 1.18 lesions per patient; p=0.01). None of our patients showed additional lesions with F-18-FDG in comparison to F-18-DOPA. Patient based detection rate increased significantly with Ct levels >= 150 pg/mL vs. Ct<150 pg/mL for both F-18-DOPA (sensitivity 90.9% vs. 28.6%; p=0.013) and F-18-FDG PET/CT (sensitivity 72.7% vs. 14.3%; p=0.025). Using a CEA cutoff of >= 5 ng/mL, detection rates of F-18-DOPA and F-18-FDG PET/CT were 81.1% and 72.7%, respectively. No correlation between Ct-DT or CEA-DT and PET positivity was found. Histological confirmation was obtained in eight patients. Conclusions F-18-DOPA PET/CT appears to be superior to F-18-FDG PET/CT in detecting and locating lesions in patients with recurrent MTC. This technique tends to be especially useful in patients with negative results in other imaging modalities and Ct >= 150 pg/mL or CEA >= 5 ng/mL.