Objective: Standard uptake values (SUVs) are widely used for quantifying the uptake of 18F-fluorodeoxyglucose (18F-FDG) in tumours. The objective of this study was to evaluate the accuracy of SUVs for malignancy in lung nodules/masses and to analyse the effects of tumour size, blood glucose levels and different body weight corrections on SUV.
Methods: One hundred and twenty-seven patients with suspicious lung lesions imaged with 18F-FDG positron emission tomography (PET) were studied retrospectively. Pathology results were used to establish lesion diagnosis in all cases. SUVs based on maximum pixel values were obtained by placing regions of interest around the focus of abnormal 18F-FDG uptake in the lungs. The SUVs were calculated using the following normalizations: body weight (BW), lean body weight (LBW), scaled body surface area (BSA), blood glucose level (Glu) and tumour size (Tsize). Receivers operating characteristic (ROC) curves were generated to compare the accuracy of different methods of SUV calculation.
Results: The areas under the ROC curves for SUV(BW), SUV(BW+Glu), SUV(LBW), SUV(LBW+Glu), SUV(BSA), SUV(BSA+Glu) and SUV(BW+Tsize) were 0.915, 0.912, 0.911, 0.912, 0.916, 0.909 and 0.864, respectively.
Conclusion: The accuracy of SUV analysis for malignancy in lung nodules/masses is not improved by correction for blood glucose or tumour size or by normalizing for body surface area or lean body weight instead of body weight.