Dynamic PET with FDG in patients with unresectable aggressive fibromatosis: regression-based parametric images and correlation to the FDG kinetics based on a 2-tissue compartment model.

OBJECTIVE: Dynamic PET (dPET) studies with 18F-FDG were performed in patients with unresectable aggressive fibromatosis before imatinib therapy. The goal of the study was to evaluate the impact of regression-based parametric imaging on tumor diagnostics. A comparison between the regression-based quantitative data (slope and intercept values) with the compartmental data of FDG was performed.
METHODS: The evaluation includes 24 patients with recurrent disease (n = 14), residual tissue (n = 2), or primary disease (n = 8), who were scheduled for palliative treatment with imatinib. Parametric images were calculated based on the dPET data by fitting a linear regression function to the time-activity data and for each voxel. Images of the slope and the intercept of the time-activity data were calculated using a dedicated software. A volume-of-interest-based analysis was also performed by applying a 2-tissue compartment model to the dPET data. The resulting parameters of the FDG kinetics [blood volume (VB), k1-k4] were compared with the volume-of-interest-based slope and intercept data. The evaluation of the parametric images was performed visually and quantitatively.
RESULTS: Twenty of 24 tumors could be visualized in the SUV images with a moderate uptake, in locations that were already known from the MR images. Most (16/24) of the tumors demonstrated a clear enhancement in the intercept images, whereas 4 of them showed an intermediate enhancement and only 4 did not show any enhancement in the intercept images. In contrast, only 10 of 24 tumors demonstrated a clearly enhanced slope, 3 of them revealed a slightly enhanced slope, and 11 of the 24 patients did not demonstrate any slope enhancement within the area of the known desmoid tumors. The comparison of slope revealed the highest correlation to the SUV (r = 0.56, P < 0.05), whereas the intercept values demonstrated the highest correlation to k1 (r = 0.794, P < 0.05), followed by the fractional VB (r = 0.709, P < 0.05), followed by SUV (r = 0.630, P < 0.05). The results indicate that slope images are related to the transport/phosphorylation-dependent part of FDG, whereas intercept images are related to the transport/perfusion part of FDG.
CONCLUSIONS: These data demonstrate that the use of regression-based parametric imaging helps to differentiate between transport/perfusion- and transport/phosphorylation-dependent FDG uptake and demonstrate that the transport/phosphorylation rate is low in most of these tumors.