Standardized uptake values (SUVs) are widely used to measure 18F-fluorodeoxyglucose (FDG) uptake in various tumors. It has been reported that normalization of FDG uptake for patient body weight (SUVbw) overestimates FDG uptake in heavy patients, as their fraction of body fat (with low FDG uptake) is often increased. The objective of this study was to determine if "normalization of FDG uptake for the body surface area" (SUVbsa) is independent of the patient's body size and is more reliable than SUVbw.
Methods: FDG-PET images were acquired on 44 patients (body weight range: 45-115 kg) with cancer. SUVbw [(mCi/g of tissue)/(mCi injected/patient body weight in g)] and SUVbsa [(mCi/g of tissue)/(mCi injected/patient BSA in m2)] were determined for the liver. Since most observers are accustomed to using the SUVbw, the two values were compared by setting the mean SUVbsa equal to that of SUVbw.
Results: SUVbw and SUVbsa were 3.42 +/- 0.85 (mean +/- s.d.) and 3.42 +/- 0.60, respectively. The standard deviation of the SUVbsa was smaller than that of SUVbw. More importantly, there was a strong positive correlation between SUVbw and, not only body weight (r = 0.75) but also BSA (r = 0.68), whereas only a weak correlation between SUVbsa and body weight (r = 0.41) or BSA (r = 0.38) was found with a near flat regression line.
Conclusion: SUVbw overestimates FDG uptake in large patients. SUVbsa appears preferable to SUVbw, since it is minimally affected by the body size.