A mathematical model for the distribution of fluorodeoxyglucose in humans

The goals of this study were to define the total body distribution kinetics of F-18-fluorodeoxyglucose (FDG), to contribute to its radiation dosimetry and to define a suitable proxy for arterial cannulation in human FDG studi es. Methods: Time-activity FDG heart, lung, liver and blood data from paire d fasting and glucose-loaded sessions in five adult human volunteers, toget her with published brain parameters, were incorporated into a multicompartm ental model for whole-body FDG kinetics. Tau values were calculated from th is model. We also compared the usefulness of activity in the left ventricle (LV), right ventricle (RV), left lung and right lung as proxy for arterial blood FDG sampling. Results: No systematic difference was found in model p arameters between the fasting and glucose-fed sessions, even for the parame ter for transfer of FDG into the myocardium. Myocardial PET data fitted wel l to a model in which there is very rapid exchange indistinguishable from b lood kinetics and transfer into an intracellular "sink." The lung data fitt ed to a simple sink representing the lung cells. The liver data required an additional intermediate exchange compartment between the plasma and a hepa tic sink. In terms of total body distribution kinetics, unmeasured organs a nd tissues (probably the skeletal muscle and gut) become increasingly impor tant with time and account for a mean of 76% of the decay-corrected FDG act ivity at infinity. Right lung activity, corrected to venous blood, represen ts the whole arterial blood curve better than the LV or RV. The tau values for radiation dosimetry of FDG in the heart, lungs, liver and bladder calcu lated from our model do not differ significantly from published results usi ng other methods. Bladder tau decreased with voiding frequency and was mark edly decreased with early voiding. Conclusion: Glucose loading state is not a good predictor of myocardial FDG uptake. The majority of FDG distributio n at 90 min is in tissues other than the blood, brain, heart and liver. Bla dder radiation will be much reduced if the patient voids early after FDG ad ministration. Summed large volume right lung activity, normalized to venous blood activity, is a good proxy for arterial blood FDG sampling. The model presented may be expanded to include other FDG kinetics as studies become available.