Abstract
Background. Non-invasive visualization of tumor biological and molecular processes of importance to diagnosis and treatment response is likely to be critical in individualized cancer therapy. Since conventional static 18F-FDG PET with calculation of the semi-quantitative parameter standardized uptake value (SUV) may be subject to many sources of variability, we here present an approach of quantifying the 18F-FDG uptake by analytic two-tissue compartment modeling, extracting kinetic tumor parameters from dynamic 18F-FDG PET. Further, we evaluate the potential of such parameters in radiotherapy response assessment. Material and methods. Male, athymic mice with prostate carcinoma xenografts were subjected to dynamic PET either untreated (n=8) or 24 h post-irradiation (7.5 Gy single dose, n=8). After 10 h of fasting, intravenous bolus injections of 10–15 MBq 18F-FDG were administered and a 1 h dynamic PET scan was performed. 4D emission data were reconstructed using OSEM-MAP, before remote post-processing. Individual arterial input functions were extracted from the image series. Subsequently, tumor 18F-FDG uptake was fitted voxel-by-voxel to a compartment model, producing kinetic parameter maps. Results. The kinetic model separated the 18F-FDG uptake into free and bound tracer and quantified three parameters; forward tracer diffusion (k1), backward tracer diffusion (k2), and rate of 18F-FDG phosphorylation, i.e. the glucose metabolism (k3). The fitted kinetic model gave a goodness of fit (r2) to the observed data ranging from 0.91 to 0.99, and produced parametrical images of all tumors included in the study. Untreated tumors showed homogeneous intra-group median values of all three parameters (k1, k2 and k3), whereas the parameters significantly increased in the tumors irradiated 24 h prior to 18F-FDG PET. Conclusions. This study demonstrates the feasibility of a two-tissue compartment kinetic analysis of dynamic 18F-FDG PET images. If validated, extracted parametrical maps might contribute to tumor biological characterization and radiotherapy response assessment.
Acknowledgements
The authors thank Professor F. Saatcioglu at Department of Molecular Biosciences, University of Oslo, for providing the CWR22 xenograft model. Financial support received from the South-Eastern Norway Regional Health Authority (grant 2009070 to K. Røe and grant 2010079 to L. B. Nilsen) and the Norwegian Cancer Society (grant 80114001 to T. Seierstad).
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.