MPTH-16. EXPRESSION OF METABOLIC ENZYMES AS PREDICTORS OF PROGNOSIS IN PATIENTS WITH AGGRESSIVE PRIMARY BRAIN CANCERS

Abstract

Aggressive primary brain cancers like glioblastoma multiforme (GBM) are metabolically active tumors. We hypothesis that abnormal expression of metabolic enzymes may reflect the tumors increased metabolism and be useful in predicting survival. 60 patients (45 with GBM, 11 with anaplastic gliomas and 4 with gliosarcoma) were evaluated using Immunohistochemistry (IHC) to measure the relative concentrations of the glycolytic enzyme hexokinase-II (HK2) a known predictor of poor prognosis and several ketolytic enzymes including cytosolic type 2(R)-hydroxybutyrate dehydrogenase (BDH-2). The immunoreactivities were graded based on the percentage of positive cells: Positive (≥20%), Negative (< 20%). Survival was calculated using Kaplan Meier (KM) and an age adjusted hazard ratio (aHR) for death was obtained using Cox proportional analysis for enzyme expression. Significance p ≤ 0.05. Demographics for 60 patients; 19 female and 41 males, median age 62 years, and median survival for all the patients 5.9 months. Univariate analysis showed positive HK2 and negative BDH-2 expression had decreased survival with HR of 1.8 (p=0.06) and 2.6 (p=0.01) respectively. In a multivariate analysis after adjusting to age, histology and other enzyme expression, positive HK2 and negative BDH-2 expression were statistically associated with decreased survival (aHR of 11.3 p= 0.001 and 32.6, p=0.0004), respectively. The group with favorable prognosis (negative HK2 or positive BDH-2 expression) (n=36) had a median survival of 9.7 months (95% CI 4.2 -13.5) whereas the group with non-favorable prognosis (n=24) had a median survival of 3.1months (95% CI 1.8 – 5.8). IHC helps to confirm the poor prognosis associated with positive HK-2 and suggests that negative IHC expression of BDH-2 may be a new indicator of poor prognosis for patients with aggressive primary brain cancers.