J Korean Radiol Soc.  1994 Sep;31(3):377-384.

Tumor Grading of Adult Astrocytic Glioma on MR Imaging

• Byun HS
• Suh DC
• Choi KH
• Han MH
• Choi CG
• Chang KH
• Suh JH
• Choi WS
• Lee SK
• Lee HK

Abstract

PURPOSE
The purpose of this study is to determine predictive MR features for grading of astrocytic gliomas and to evaluate the accuracy of MR grading in these tumors.
MATERIALS AND METHODS
We retrospectively reviewed 135 cases of supratentorial astrocytic gliomas in adult (age > 15 years), all of which were proved by open biopsy. Two observers analysed MR images independently with criteria of margin, edema, mass effect, signal heterogeneity, necrcosis, cyst formation, hemorrhage, tumor vascularity, enhancement degree, and enhancement size. The patterns of enhancement were categorized into no, homogeneous, heterogeneous, thin smooth rim, thin irregular rim, and thick irregular rim enhancement patterns. Observers finally diagnosed each case as one of low-grade astrocytoma, anaplastic astrocytoma or glioblastoma multiforme.
RESULTS
Statistically significant MR features for grading of these tumors were revealed as necrosis (p < 0. 001), edema (0.008), and signal heterogeneity (p < 0.025). When compared with histopathologic grading, MR graded correctly 76%-77% of cases in two tiered system(low-grade astrocytoma versus high-grade astrocytoma), but only 67%-69% of cases in three tiered system(low-grade astrocytoma, anaplastic astrocytoma, glioblastoma multiforme). No contrast enhancement was seen in 45% and 23% of low-grade astrocytoma and anaplastic astrocytoma respectively. Glioblastoma multiforme frequently showed thick irregular rim enhancement (57%), but no enhancement at all in 8%.
CONCLUSION
We have concluded that necrosis and edema are significant predictive MR features for gradomg of supratentorial astrocytic gliomas in adult, and MR was correct in 76%--77% of cases for predicting pathologic grading of astrocytomas in two tiered system.

MeSH Terms

Adult*
Astrocytoma*
Biopsy
Edema
Glioblastoma
Hemorrhage
Humans
Magnetic Resonance Imaging*
Necrosis
Neoplasm Grading*
Population Characteristics
Retrospective Studies