Pathologic diagnosis of recurrent glioblastoma: morphologic, immunohistochemical, and molecular analysis of 20 paired cases.

To evaluate the prognostic value of the volume of residual viable tumor versus therapy-induced necrosis in resection material and the diagnostic value of ancillary tests in recurrent glioblastoma (GBM), we conducted a retrospective review of 20 patients whose initial and recurrent specimens were available. Recurrent GBMs were graded according to the extent of histopathologic parameters: recurrent tumor with high-grade, non-high-grade, and pure high-grade tumor components and therapy-related necrosis. We also examined MIB-1 labeling, isocitrate dehydrogenase 1 mutation, and epidermal growth factor receptor amplification in primary and recurrent GBMs. To evaluate patient outcomes according to clinical and pathologic parameters, a survival analysis was performed, and correlations between histopathologic parameters and each ancillary test were assessed. Among clinical parameters, age above 60 years was associated with decreased survival (P=0.022), but other clinical parameters showed no significant association with overall survival. Among the 3 histopathologic parameters, the extent of recurrent tumor, including high-grade and non-high-grade components, revealed a significant association with overall survival (P=0.042), but neither the extent of pure high-grade components nor therapy-related necrosis showed any prognostic value. MIB-1 labeling, isocitrate dehydrogenase 1 mutation, and epidermal growth factor receptor amplification were useful for the diagnosis of recurrent GBMs but showed no prognostic value. Our data suggest that histopathologic evaluation on the basis of tumor extent in resected recurrent GBM specimens may provide additional prognostic information on the survival of patients with recurrent GBM.