Monitoring brain tumor therapy with 18F-FLT

Abstract

16

Objectives: FLT is a predictor of survival in recurrent glioma therapy (JCO 2007;25:4714-21). We investigated changes in FLT kinetics during treatment.

Methods: High-grade brain tumors were investigated in thirteen patients (5 men, 8 women), 26-70 yr. Each had 3 dynamic PET studies, at baseline (study 1), 2 weeks after start of treatment (study 2) and 6 weeks (study 3). 1.5 MBq/kg 18F-FLT was administered, dynamic images acquired for 1 hr and iteratively reconstructed. PET studies were coregistered and resampled with SPM to contrast enhanced baseline T1-weighted MR with isotropic voxels. Factor analysis generated factor images, from which blood curve and tumor uptake curves were derived. A 75% isocontour of the tumor maximum on the tumor factor image was used to define ROIs for each study. A cubic ROI of 1.8 mL was also defined on the baseline scan and the 3 studies processed with the ROI in the same position. A 3-compartment, 2-tissue model was applied with metabolite and partial volume corrections to estimate the rate constants.

Results: No significant differences were found between parameters derived from the original vs the MR-aligned PET images (p>0.1; r >0.9). A linear relationship was found between transport rate k1 and influx rate K, and between K and SUV. Significant differences (p<0.05) were found for K and SUV between all 3 studies, and for k1 in study 1 vs 2. From study 1 to 2, K decreased by 26%, SUV by 30% and k1 by 36%. During 2 weeks follow up, the largest change in k1, K, and SUV was seen in patients with a favorable response. Parameter changes are correlated with patient survival.

Conclusions: Compartmental analysis of FLT during treatment shows a linear relation between K and SUV, of which the sharpest decline is between baseline and 2 weeks follow up. It appears adequate to monitor therapy response with SUV in clinical practice.