Background: Glioblastoma is relatively uncommon in childhood and maybe difficult to differentiate from other brain tumors such as primitive neuroectodermal tumor, ependymoma, or benign astrocytoma.
Objective: To describe the characteristic MR features in children with glioblastoma and to evaluate the usefulness of diffusion and perfusion MR imaging and MR spectroscopy in pediatric glioblastoma.
Materials and methods: MR imaging in 11 children (12 tumors) with biopsy-proven glioblastoma was reviewed retrospectively. In one patient, there was a recurrent glioblastoma. We reviewed CT and MRI imaging for tumor location, density/signal intensity, and enhancement pattern. Routine MR imaging was performed with a 1.5-T scanner. In six patients, diffusion-weighted MR images (DWIs) were obtained with a single-shot spin echo EPI technique with two gradient steps, and apparent diffusion coefficients (ADCs) were calculated. Using the gradient EPI technique, perfusion-weighted MR images (PWIs) were obtained in four patients from the data of dynamic MR images. The maximum relative cerebral blood volume (rCBV) ratio was calculated between the tumor and contralateral white matter in two cases. In three patients, proton MR spectroscopy was performed using a single voxel technique with either STEAM or PRESS sequences. The locations of the tumor were the thalamus and basal ganglia ( n=8), deep white matter ( n=3), and brain stem ( n=1).
Results: Intratumoral hemorrhage was seen in four tumors. The tumors showed high-signal intensity or DWIs, having a wide range of ADC values of 0.53-1.30 (mean +/-SD=1.011+/-0.29). The maximum rCBV ratios of glioblastoma were 10.2 and 8.5 in two cases. MR spectroscopy showed decreased N-acetylaspartate (NAA) and increased choline in three cases. The MR findings of glioblastoma in children were: a diffusely infiltrative mass with hemorrhage involving the deep cerebral white matter, thalami, and basal ganglia.
Conclusion: Diffusion/perfusion MR imaging and MR spectroscopy are very helpful in diagnosing glioblastoma, determining the biopsy site, and evaluating tumor recurrence.