RÉSUMÉ
Objective:To investigate the diagnostic value of integrated 18F-fluorodeoxyglucose (FDG) PET/MRI multimodal imaging in the grading of adult diffuse gliomas. Methods:A retrospective analysis was performed; 29 patients accepted cranial 18F-FDG PET/MRI multimodal imaging and diagnosed as having adult diffuse gliomas in our center from January 2020 to December 2021 were chosen; their clinical and image data were collected. These patients were divided into low-grade group (WHO grading II, n=13) and high-grade group (WHO grading III and IV, n=16) according to the pathological results confirmed by surgery or biopsy. T1WI, T2WI, DWI, arterial spin labeling (ASL), magnetic resonance spectroscopy (MRS) and PET were performed in these patients. Differences in T1WI and T2WI signs between the two groups were compared. Differences in maximum standard uptake value (SUV max), apparent diffusion coefficient (ADC), choline/ N-acetylaspartate (Cho/NAA), NAA/creatine (Cr), Cho/Cr, relative cerebral blood flow (rCBF) between the two groups were compared. The diagnostic efficacies of above parameters in differentiating high-grade from low-grade adult diffuse gliomas were determined by receiver-operating characteristic (ROC) curves. The predictive probabilities for differentiating high-grade from low-grade adult diffuse gliomas using combined parameters were obtained by binary Logistic regression analysis. Results:There were no statistical differences in T1WI and T2WI signs, proportion of adjacent parenchyma edema, proportion of crossing the midline into the contralateral side, proportion of mass effect, proportion of cystic degeneration/necrosis, or site of growth between low-grade and high-grade groups ( P>0.05). The values of SUV max, ADC, Cho/NAA, Cho/Cr and rCBF were statistically different between the two groups, and the order of diagnostic efficiency was SUV max>ADC>Cho/NAA>rCBF>Cho/Cr>NAA/Cr. SUV max had the highest diagnostic efficacy in differentiating high-grade from low-grade adult diffuse gliomas with an area under the curve (AUC) of 0.894, a sensitivity of 75% and a specificity of 100%; the diagnostic efficacy of combined parameters (mean AUC>0.9) was higher than that of single parameter, and the combined parameters of SUV max and Cho/NAA showed the highest diagnostic efficacy with an AUC of 0.933, a sensitivity of 81.3% and a specificity of 92.3%. Conclusion:The 18F-FDG PET/MRI multimodal imaging is helpful in the grading diagnosis of adult diffuse gliomas, and combination of PET and MRS may play a greater role.