ASSESSMENT OF TUMOR HYPOXIA AND PERFUSION IN GBM FOLLOWING BEV FAILURE USING FMISO 18F-PET AND MRI

Glioblastoma (GBM, WHO grade IV) is the most common and aggressive of the primary brain tumors in adults. As part of an ongoing clinical trial with a hypoxia activated prodrug, we sought to explore the correlation between abnormal tumor vasculature and hypoxia in GBM following bevacizumab (BEV) failure. MRI and 18F-FMISO PET scans were acquired at study entry. Associations between baseline hypoxia volume (HV), SUV peak, as measured by FMISO PET and vascular parameters including rCBV, TTP, enhancing and non-enhancing tumor volumes as measured by dynamic susceptibility contrast imaging (DSC), T1 post contrast and FLAIR was determined using Pearson correlation. Each marker was modeled with a univariate cox regression model for progression free survival (PFS) time. 8 patients from University of Texas Health Science center and 14 patients from Dana-Farber Cancer Institute had evaluable MR and PET imaging. Significant positive correlations were found between HV and enhancing volume (R=0.6884, p<0.0001) as well as volume ratio and srCBV (R=0.6032, p<0.0001). Moderate positive correlation were found between volume ratio and HV (R=0.5049, p=0.0165) as well as volume ratio and SUVpeak (R=0.5669, p=0.0059). Nonenhancing volume is associated with HV (R=0.2602, p=0.0153) so is SUVpeak and srCBV (R=0.3253, p=0.0056). In the univariate cox model, bigger enhancing (p=0.043), nonenhancing (p=0.0061) and HV (p=0.0172) were significantly associated with shorter progression time. The hypoxic volume following bevacizumab failure correlates with both the volume of enhancement and the fraction of enhancement within the mass. Findings from the full 23 patient cohorts and the association with other biomarkers and response to treatment is being assessed and will be presented.