Total lesion glycolysis as the best 18F-FDG PET/CT parameter in differentiating intermediate–high risk adrenal incidentaloma

다국어 초록 (Multilingual Abstract)

OBJECTIVE: Characterization of intermediate–high risk adrenal incidentaloma (AI) is important because biopsy or surgery should be performed to confirm the malignancy. We investigated which parameters of F-fluorodeoxyglucose (F-FDG) PET/computed tomography (CT) had an additive role in distinguishing malignancies in patients with incidental adrenal masses of intermediate–high risk. METHODS: From January 2008 to July 2013, 52 patients with a pathologically proven diagnosis of AI were retrospectively enrolled (age=56.4±12.7 years, M : F=34 : 18; benign : malignant=14 : 38). Volumetric parameters were size and volume according to combined CT, and metabolic parameters were peak standardized uptake value (SUVpeak), maximum SUV (SUVmax), mean SUV (SUVmean), and tumor-to-background ratio (SUVmax of adrenal mass/SUVmean of liver). Metabolovolumetric parameters of metabolic tumor volume and total lesion glycolysis (TLG, SUVmean×metabolic tumor volume) were also included and compared with the diagnostic value. In addition, the highest diagnostic parameters among volumetric and metabolic parameters were combined and compared in terms of diagnostic accuracy. RESULTS: Compared with benign adrenal adenoma, malignant lesions showed significantly higher values of all F-FDG PET/CT volumetric, metabolic, and metabolovolumetric parameters. Size showed the highest area under the curve (AUC) of 0.759 among the volumetric parameters, and SUVpeak showed the highest AUC of 0.853 among the metabolic parameters. Among all the PET/CT parameters, TLG showed the highest AUC of 0.900, with a sensitivity of 92.1% and specificity of 78.6% at a cutoff of 12.0. The combined value of size and SUVpeak showed lower diagnostic value than TLG. CONCLUSION: We found that TLG showed the best result in distinguishing intermediate–high risk AI among PET/CT parameters. TLG can be a useful PET/CT parameter for differential diagnosis of AI.