Abstract
370
Objectives PET imaging and early detection of lesions is critical for disease staging and treatment decisions. Whereas standard clinical 18F-FDG-PET imaging time is 60 minutes post radiotracer injection, studies have indicated that imaging at a later time can improve detectability as a result of continuing uptake in malignant tumors [1,2,3]. The 'untrapping' rate (k4) of 18F-FDG from breast cancers is small and has not been well quantified [1,4]. While the effects are negligible for imaging at 60 minutes post injection, we wanted to determine the impact of a non-zero rate on detectability for longer uptake periods.
Methods We combined a kinetic model of 18F-FDG radiotracer uptake [5,6], a detailed forward model of PET data acquisition [7] and reconstruction, and observer models to estimate lesion detectability [8,9]. De-phosphorylation rates of k4 = 0, 0.001, 0.005 and 0.01 were used. Kinetic parameters were used to generate time activity curves out to 10 hours. We generated 600 i.i.d. noisy sinogram realizations for 5 time-points. Images were reconstructed using post-filtered OSEM. Both non-prewhitening and channelized Hoteling model observer statistics were calculated.
Results Detectability increased for several hours past the current standard post-injection imaging window of one hour, although the optimum time shifted to earlier time points with increasing k4. This improvement is driven by the continued accumulation of FDG in the tumor before the impact of increased noise due to radiotracer decay begins to dominate.
Conclusions These results imply that delayed PET imaging may reveal low-conspicuity lesions in patients that would have otherwise gone undetected. Future studies may also address patient prognosis with this factor, taking into account other clinically relevant variables known to be important in the context of breast cancer outcome [10].
Research Support NIH 1RC1CA146456, CA148131, CA72064, CA42045, CA138293