Simultaneous [18F]FET PET/MR in cerebral tumors. Comparison with conventional PET/CT

256 Objectives To compare diagnostic performance between PET/CT and recently introduced intergrated whole-body PET/MR concerning evaluation of brain tumours. Methods Twenty pts, referred to [18F]FET-PET for evaluation of brain tumors, underwent a single-injection/dual-imaging protocol, including a PET/CT scan (Siemens Truepoint 64, 32±5 min p.i., 10 min 3D acquisition) and a subsequent MR/PET scan (Siemens mMR, 52±8 min p.i., 15 min 3D acquisition), after injection of 183±44 MBq [18F]FET. Images were reconstructed with filtered backprojection (Hann 4.9mm, zoom 2.5), attenuation correction was performed using low dose CT for the PET/CT and Dixon-MR sequences for the PET/MR. Images were interpreted visually and by semiquantitative analysis. The respective lesion and Bg SUVs were compared as well as the lesion/background ratios (T/Bg) using the wilcoxon test. Results Visual interpretation turned out suspicion of tumour tissue in 16/20 scans with both PET/MR and PET/CT, 4/20 patients were read as negative. Overall, SUVmax, SUVmean of the lesion and SUVmean of the Bg. calculated in the respective scans were significantly different: Average lesion SUVmax PET/CT: 3.1±1.2, average lesion SUVmax PET/MR: 2.7±1.0; av. lesion SUVmean PET/CT: 2.7±1.1, av. lesion SUVmean PET/MR: 2.2±0.8; av. Bg SUVmean PET/CT: 1.2±0.2, av. Bg SUVmean PET/MR: 1.0±0.2 (p Conclusions Imaging of brain tumors with [18F]FET is feasible with the integrated whole-body PET/MR-scanner with high image quality. Despite technological differences, the semiquantitative evaluation of [18F]FET-PET data with PET/MR concerning T/Bg. ratios reveals comparable results as with PET/CT, indicating that established lesion/background ratio thresholds for distinguishing malignant brain tumor tissue, can be transferred