Value of MRI in grading of supratentorial diffuse astrocytomas

ABSTRACT

The present study included 15 patients [8 males and 7 females] with their age ranged from 25 years to 68 years. The MR images in the studied patients with biopsy-proven diffuse ostrocytomas were analyzed in order to demonstrate which MRI features of gliomas that correlate with the histopathologically determined tumor grade. According to the histopathological diagnosis, 3 patients had grade II astrocytomas, 5 patients had grade III astrocytomas and 7 patients had glioblastoma multiformes [GBMs]. The MRI features studied included tumor heterogeneity, edema, mass effect, border sharpness, contrast enhancement, hemorrhage and the presence of flow voids. Contrast enhancement was found to be the best predictor of histological grade followed by necrosis, signal heterogeneity and border sharpness. Hemorrhage, flow void and necrosis were common in the GBM group and were absent in grade II astrocytoma group. In particular, flow void was found only in GBM [n=3 patients], hemorrhage was present in GBM [n= 3 patients] and in grade III astrocytoma [one patient]. Necrosis was found in all 7 GBM cases and in 5 patients with histological diagnosis of anaplastic astrocytoma [grade III astrocytoma]. No areas of necrosis detected in grade II astrocytoma [3 patients]. All GBM cases [n = 7] showed grade 3 edema [larger than the tumor volume], marked and heterogeneous enhancement, midline shift and heterogeneous signal intensity on both Tl-W and T2-W images. In grade III astrocytomas [n = 5 patients], there is edema of grade 2 [less than the tumor volume], moderate contrast enhancement, midline shift and heterogeneous signal intensity on T2-W images. In grade II astrocytomas [n = 3 patients], there is edema of grade 1 [none-to mild], subarachnoid space effacement, none- to mild contrast enhancement and homogeneous signal intensity. The frontal lobe was the most commonly affected lobe, frontal gliomas were present in 9 patients, fronto-parietal gliomas in 4 patients and fronto-temporal gliomas in 2 patients. In conclusion, all MRI features can help to distinguish the pathological grade of diffuse astrocytic tumours and help in planning treatment strategies