Abstract
Identifying predictive and prognostic markers for glioblastoma (GBM) is an area of intense investigation. The objective of this study was to evaluate OLIG2 and Ki-67 as prognostic biomarkers. A tissue microarray with clinical data was created for 70 patients diagnosed with GBM between 2002-2007 at a single institution. Immunohistochemistry was performed to evaluate expression of OLIG2 and Ki-67. Overall survival and progression free survival (PFS) were calculated using the Kaplan-Meier method. Univariable (UVA) and multivariable (MVA) analysis was performed using Cox proportional hazards modeling. The median age of diagnosis was 63, 66% of patients were male, 80% had a KPS ≥70, 64% had a subtotal resection or biopsy, and 36% had a gross total resection. Median OS and PFS were 13.3 months and 9.5 months respectively. For patients who received resection plus chemoradiation, median OS and PFS were 16.7 months and 9.5 months respectively. 36.4% of patient samples had Ki-67 expression fraction > 20%. 18.0% had heavy OLIG2 staining. On MVA, lower KPS was associated with shorter OS and radiation therapy was associated with longer OS. OLIG2 was not associated with OS or PFS. Patients who had a Ki-67 < 20% had a shorter PFS, but not OS on MVA. Our data demonstrate higher Ki-67 levels correlate with improved PFS without an OS benefit. This association persisted for patients receiving standard of care treatment. OLIG2 levels did not correlate with PFS or OS in either cohort. Our study indicates that further investigation is needed to determine whether Ki-67 is useful in predicting recurrence risk and potentially changing treatment strategies for these patients, as well as the mechanism behind the associated improvement seen with higher Ki-67 levels. These data are currently limited by lack of IDH and MGMT status but suggest Ki-67 should be further investigated as a biomarker for progression.