P08.14 In situ detection of hypoxia inducible factor 2 alpha in malignant gliomas

Abstract

INTRODUCTION: Direct intratumoral measurements confirm that necrosis and hypoxia are fundamental features of the pathobiology of glioblastoma (GBM). Tumor cells adapt to hypoxia through up regulation of hypoxia inducible factors (HIFs), a conserved family of transcription factors. Of the HIFs, HIF-2 alpha (HIF-2α) appears to be expressed in GBM tumor cells (not neural progenitors) and drives adaptation to prolonged (>24 hour) hypoxia. Targeting HIFs is an attractive cancer treatment strategy given that tumor hypoxia is a constant environmental cue driving genetic instability, cell migration, angiogenesis, and chemoradiation resistance. GBM, a tumor uniformly characterized by an intratumoral hypoxic environment, is a prime candidate for HIF therapeutic targeting but specific drugs were not available. PT2385 is a novel, first-in-class, HIF-2α inhibitor which recently entered clinical trials for renal carcinoma. PT2385 is orally administered and is small, lipophilic with a brain:plasma ratio of 0.9 in rats. To establish HIF-2α as a therapeutic target in GBM, we investigated in situ expression of HIF-2α protein in tissue samples from the Wake Forest Brain Tumor Center of Excellence Brain Tumor Bank. METHODS: 22 formalin-fixed, paraffin-embedded glioma samples (grade II-IV) were analyzed for HIF-2α expression using immunohistochemistry. After rehydration, endogenous peroxidase was blocked, epitopes were retrieved using a microwaved Tris EDTA pH 9.0 solution, non-specific epitopes were blocked, and HIF-2α antibody (Santa Cruz, sc-13596) was added. A horseradish peroxidase conjugated anti-mouse antibody was then added followed by detection chromagen. An inverted bright field microscope captured images and representative samples were digitally scanned. Localization and quantification of HIF-2α was independently verified by a neuropathologist. RESULTS: There was no detectable HIF-2α expression in the four Grade II and two Grade III gliomas studied. Of the 16 GBMs (Grade IV), HIF-2α was expressed in 13 (81%). HIF-2α was highly expressed in seven specimens (>10% cells positive), intermediate in six specimens (<10% of cells positive), and minimal to none in three specimens. Staining was specific to both tumor cells and occasionally monocytic cells (based on morphology). It was noted that HIF-2α was frequently present in perivascular and perinecrotic regions. CONCLUSIONS: HIF-2α expression is found in the majority of GBM specimens, but is absent in low-grade gliomas. Immunohistochemistry demonstrated a range of HIF-2α abundance along with regional staining patterns, clustered in perivascular and perinecrotic niches. HIF-2α appears to correlate with increasing malignancy grade. This is the first in situ description of HIF-2α in gliomas and further studies are in progress for preclinical in vitro and in vivo testing of PT2385, a first-in-class HIF-2α targeted agent, in GBM.