Abstract
OBJECTIVES: To investigate the effects of ambient glucose on quantitative analysis of hepatic tumors on 2-deoxy-2-((18)F)-fluoro-D-glucose ((18)FDG) positron emission tomography (PET) and to establish a method for glucose correction. PATIENTS AND METHODS: Eighty-six patients with hepatic lesions identified on (18)FDG PET/computed tomography (CT) were analyzed. The serum glucose level (Glc) was recorded prior to imaging, and the maximum standardized uptake value (SUV) in the hepatic tumors and the average SUV in normal liver were determined. The inverse relationship of SUV to glucose can be defined as d (SUV)/d (Glc) = g*SUV/(Glc), where g is the glucose sensitivity. Simulations using glucose level from 70 to 250 mg/dl were performed to evaluate the effects of Glc on the maximum SUV of malignant hepatic lesions and normal liver. RESULTS: By logarithmic transformation and linear regression, g for metastasis was significantly higher than that for normal liver (-0.636 ± 0.144 vs. -0.0536 ± 0.0583; P = 0.00092). Simulation studies showed that the SUV in malignant lesions will decrease rapidly when Glc level is >120 mg/dl, while background liver remains relatively constant up to 250 mg/dl. CONCLUSION: The tumor FDG uptake is much more sensitive to ambient glucose level variation than the background liver. Therefore, correction by the glucose sensitivity factor will result in more accurate SUV measurements and make semi-quantitative analysis of (18)FDG PET scans more reliable.