Metabolic Imaging of Cerebral Gliomas: Spatial Correlation of Changes in O-(2-18F-Fluoroethyl)-l-Tyrosine PET and Proton Magnetic Resonance Spectroscopic Imaging

The aim of this study was to determine the spatial correlation of O-(2-18F-fluoroethyl)-l-tyrosine (18F-FET) uptake and the concentrations of choline (Cho), creatine (Cr), and total N-acetylaspartate (tNAA) determined with proton magnetic resonance spectroscopic imaging (1H MRSI) in cerebral gliomas for the multimodal evaluation of metabolic changes. Methods: 18F-FET PET and 2-dimensional 1H MRSI were performed in 15 patients with cerebral gliomas of World Health Organization (WHO) grades II–IV. PET and 1H MRSI datasets were coregistered by use of mutual information. On the basis of their levels of 18F-FET uptake, 4 different areas in a tumor (maximum, strong, moderate, and low 18F-FET uptake) were defined on PET slices as being congruent with the volume of interest in the 1H MRSI experiment. 18F-FET uptake in lesions was evaluated as tumor-to-brain ratios. Metabolite concentrations for Cho, Cr, and tNAA and Cho/tNAA ratios were computed for these 4 areas in the tumor and for the contralateral normal brain. Results: In the area with maximum 18F-FET uptake, the concentration of tNAA (R = −0.588) and the Cho/tNAA ratio (R = 0.945) correlated significantly with 18F-FET uptake. In the areas with strong and moderate 18F-FET uptake, only the Cho/tNAA ratios (R = 0.811 and R = 0.531, respectively) were significantly associated with amino acid transport. At low 18F-FET uptake, analysis of the correlations of amino acid uptake and metabolite concentrations yielded a significant result only for the concentration of Cr (R = 0.626). No correlation was found for metabolite concentrations determined with 1H MRSI and 18F-FET uptake in normal brain tissue. Maximum 18F-FET uptake and the tNAA concentration were significantly different between gliomas of WHO grades II and IV, with P values of 0.032 and 0.016, respectively. Conclusion: High 18F-FET uptake, which is indicative of tumor cell infiltration, associates with neuronal cell loss (tNAA) and changes in ratios between parameters representing membrane proliferation and those of neuronal loss (Cho/tNAA ratio), which can be measured by 1H MRSI. The significant correlation coefficients detected for Cr in regions with low 18F-FET uptake suggests an association between the mechanism governing amino acid transport and energy metabolism in areas that are infiltrated by tumor cells to a lesser extent. These findings motivate further research directed at investigating the potential of 1H MRSI to define tumor boundaries in a manner analogous to that of amino acid PET.