Multimodal hypoxia imaging in a preclinical glioma model

Introduction: Molecular imaging techniques allow early characterization of tumors and quantification of biological processes in vivo. The aim of our study is to analyze the relationship between tumor growth and tumor hypoxia in an orthotopic glioma murine model by using a multimodal procedure, and to compare the features of the different imaging techniques in revealing hypoxia induction. Materials and methods: Engineered U251 cells were gently provided by Dr. Giovanni Melillo, National Cancer Institute, Frederick (MD). These cells express the luciferase reporter gene under control of a constitutive promoter (U251-LUC) or under control of three copies of a Hypoxic Responsive Element (U251-HRE). Cells has been analyzed by means of three different approaches: 1) In vitro evaluation of transcriptional activation of HIF-1α at different times after deferoxamine (DFX) treatment by immunocytochemistry (ICC). 2) In vivo analysis of tumoral progression in orthotopic murine models obtained by stereotaxic injection of 105 glioma cells; animals were monitored weekly with BLI (CCD camera), PET ([18F]FDG,[18F]FAZA,[18F]FLT) and MRI to evaluate tumoral progression and hypoxia activation. 3) Ex vivo analysis by H&E staining and hypoxia markers (HIF-1α and CAIX) were performed. Results: In vitro studies showed no differences among the two cell lines as regards to HIF-1α activation kinetic with a peak of activation between 3 and 6h and a decrease at 20h. In vivo, U251-HRE model showed a detectable and progressive luciferase activity induction starting at 18 days from injection demonstrating that luminescence expression is dependent on HRE-mediated activation. On the other hand, in U251-LUC model luciferase activity was detectable immediately after injection and remained proportional to tumor growth. Lesions were highly proliferative ( [18F]FLT) but hypo-metabolic ([18F]FDG). In partial agreement with the results of U251-HRE, a hypoxia dependent [18F]FAZA uptake was observed at later times (30 days). MRI provided morphological characterization and diffusion studies showed hypoxic necrotic areas only at later days. Ex vivo H&E staining demonstrates that tumors were characterized by nuclear atypia, brisk mitotic activity, microvascular proliferation and necrosis. HIF-1α and CAIX were clearly expressed in hypoxic areas, especially in the inner part of the tumor. Conclusions: This study demonstrates that the U251-HRE orthotopic murine model may be proposed as a predictive and reliable tool to evaluate hypoxia dependent processes of human glioma in preclinical studies by BLI. Differences among the three imaging techniques may be related to methods sensitivity. Further ex vivo post mortem measurements of HIF-1α and CAIX at earlier times will be performed. Additional studies will be conducted to understand the clinical meaning of early or delayed hypoxia identification in terms of response to treatment. References: SEMENZA GL. Curr Opin Cell Biol 13(2):167-71, 2001. Review. RAPISARDA A. Cancer Research 62, 4316-4324, 2002