Comparison between dynamic [18F]Fluoroethyltyrosine PET/CT and advanced MRI in cerebral high and low grade gliomas

Comparison between dynamic [18F]Fluoroethyltyrosine PET/CT and advanced MRI in cerebral high and low grade gliomas L. Picori, U. Rozzanigo, D. Donner, M. Erini, P. Feraco, M. Recla, F. Chierichetti; Santa Chiara Hospital, Trento, ITALY. Aim/Introduction: To investigate if dynamic [18F]fluoroethyl- L-tyrosine [18F]FET PET/CT improves the diagnosis in patients with suspected new or recurrent cerebral gliomas, respect to advanced MRI techniques. Materials and Methods: We retrospectively evaluated 20 patients who performed [18F] FET by a PET/CT tomograph: 15 had an indeterminate brain lesion, 5 a suspect glioma recurrence. All patients underwent a 40 minutes dynamic [18F]FET PET/CT acquisition and two different sequences, between 5-15 minutes and 20-30 minutes. For dynamic studies time-activity and time to peak curves were extracted using different region-of-interest (ROIs) and volume of interest (VOIs) definitions. MRI was performed with a 1.5T scanner just before [18F]FET-PET/CT using perfusion (PWI) and diffusion (DWI) weighted imaging: afterwards rCBV and ADC values were calculated placing the VOIs on the solid components of the lesion. In case of doubt (13 cases) single-voxel MR spectroscopy was performed. Multimodality imaging by fusion of PET/CT and different MRI sequences was performed for a joint assessment (radiologist and nuclear physician). Results: Final diagnosis was based on histology in 8 patients who underwent neurosurgery (5 HGG, 3 LGG) and on follow-up imaging in 12 patients (8 tumor progression, 4 stable benign lesion). On the basis of [18F]FET- PET, 7 cases were classified as high uptake (2 glioma recurrence and 5 new diagnosis of HGG tumor), 8 as low uptake (4 glioma recurrence, 2 new diagnosis of LGG tumor, 1 tumor progression, 1 tumefactive demielinating lesion) and 5 as no uptake (1 new diagnosis of LGG tumor, 4 stable benign lesion). Sensibility for dynamic [18F]FET-PET was 93% and specificity was 80%. Multiparametric MRI was in agreement with [18F]FET-PET in all 7 cases of high uptake and in 5 cases of low uptake. [18F]FET- PET helped to classify 6 MRI indeterminate lesions (2 suspect radionecrosis with pathologic uptake, 4 benign lesions without uptake). In 2 cases there was a discrepancy between MRI and PET: 1 tumefactive demielinating lesion was classified by [18F] FET as low uptake lesion, 1 LGG confirmed at histology showed no uptake. Conclusion: In our experience, adding quantitative data, such as dynamic acquisition in PET/CT by aminoacid tracer like [18F]FET, to rCBV and ADC maps in advanced MRI is crucial for a better comprehension of tumor lesions and to assess grading. Dynamic [18F]FET-PET/CT and multiparametric MR imaging have a very high sensibility to detect new tumoral lesions or suspect glioma recurrence. Agreement between PET and MRI is essential to improve diagnostic specificity. References: None