Abstract
307
Objectives: PET-imaging with O-(2-[18F]fluoroethyl)-L-tyrosine (FET) allows not only an improved delineation of cerebral glioma, but also tumor grading based on FET kinetics. Low grade gliomas (LGG) exhibit a steadily increasing FET uptake (LG kinetics), whereas high-grade gliomas (HGG) are characterized by rapid uptake followed by washout (HG kinetics). We developed a method to characterize the FET kinetics by parametric images.
Methods: Parametric images of the y-axis intercept and the slope of a linear regression to dynamic FET data (5 to 50 min) were obtained by applying the regression analysis utility of the PMOD software. These images were corrected for global FET uptake by normalization to the mean intercept and slope of the normal cortex. Intercept x slope-product (ISP) images were calculated after all voxels with intercept values < 2 were set to zero. A first qualitative analysis was done in 10 HGG patients and 5 LGG patients. A color scale was applied presenting negative ISP as red voxels (HG kinetics) and positive ISP as green voxels (LG kinetics).
Results: In HGG patients ISP images displayed contiguous red areas of HG kinetics, but also green regions of LG kinetics with interspersed red spots of HG kinetics. The latter was seen in tumors considered to be LGG by a first ROI-based analysis, but graded as HGG histologically. In LGG, ISP images showed only green areas of LG kinetics.
Conclusions: Parametric linear regression images of FET kinetics in brain tumors facilitate the differentiation between HGG and LGG and reveal tumor heterogeneity.
- Society of Nuclear Medicine, Inc.