Usefulness of whole-body 18F-FDG PET/CT in the presurgical discrimination of cerebellar tumors

Purpose The present study aims to evaluate the value of whole-body ¹⁸ F-FDG PET/CT in distinguishing among different cerebellar tumor types. Methods We retrospectively analyzed ¹⁸ F-FDG PET/CT images of 86 patients with histologically confirmed cerebellar tumors, including 25 metastases, 17 lymphomas, 9 low-grade gliomas, 13 high-grade gliomas, 16 hemangioblastomas, and 6 medulloblastomas. Tumors were initially classified as PET-positive or PET-negative by visual assessment. For PET-positive cases, semiquantitative parameters—including the maximum, mean, and peak tumor-to-normal-brain ratios (TNRmax, TNRmean, and TNRpeak, respectively), metabolic tumor volume (MTV), and total lesion glycolysis (TLG)—were measured and compared pairwise. Parameters significant for differential diagnosis were evaluated using the area under the receiver operating characteristic curve (AUC) and accuracy. Additionally, the detection of suspicious extracranial malignancy on torso PET/CT was recorded, and its diagnostic value for metastasis was assessed. Results Nearly all hemangioblastomas were PET-negative, with visual assessment achieving a diagnostic accuracy of 97.67% for this tumor type. Lymphomas presented the highest TNRmax, TNRmean, and TNRpeak values, whereas low-grade gliomas presented the lowest values. For distinguishing lymphoma, the AUCs for TNRmax, TNRmean, and TNRpeak were 0.881, 0.889, and 0.898, respectively. When optimal cutoff values of > 1.67, > 0.94, and > 1.37 were used, the diagnostic accuracies were 71.01%, 75.36%, and 85.51%, respectively. For identifying low-grade glioma, the same parameters yielded AUCs of 0.904, 0.916, and 0.869, respectively. With optimal cutoff values of < 0.77, < 0.52, and < 0.62, the accuracies were 92.75%, 91.30%, and 89.86%, respectively. Medulloblastoma demonstrated the highest MTV and TLG. MTV yielded an AUC of 0.847 for differentiating medulloblastoma from other cerebellar tumors, and an optimal cutoff value of > 12.85 provided an accuracy of 89.86%. Torso PET/CT for detecting suspicious extracranial malignancy excelled at diagnosing metastasis, with an accuracy of 96.51%. Conclusion Visual analysis and metabolic parameters from brain ¹⁸ F-FDG PET are valuable for differentiating hemangioblastoma, lymphoma, low-grade glioma, and medulloblastoma. Whole-body PET/CT contributes to the diagnosis of metastasis by identifying suspicious extracranial malignancies.