Hypoxia Imaging with 18F-FMISO PET for Brain Tumors

Abstract

Tumor hypoxia is an important object for imaging because hypoxia is associated with tumor aggressiveness and resistance to radiation therapy. Here, 18 F-fluoromisonidazole (FMISO) has been used for many years as the most commonly employed hypoxia imaging tracer. Unlike F-18 fluorodeoxyglucose (FDG), FMISO does not accumulate in normal brain tissue making it able to provide images of hypoxic brain tumors with high contrast. Clinical evidence has suggested that FMISO PET can predict patient prognosis and treatment response. Among gliomas of various grades (WHO 2007), it has been known that grade IV glioblastoma resides under severe hypoxia and is a cause of development of necrosis in the tumor. For this study we tested whether FMISO can distinguish the oxygen condi-tion of glioblastomas and lower-grade gliomas. Twenty-three glioma patients underwent FMISO PET for the study. All the glioblastoma patients (N ¼ 14) showed high FMISO uptakes in the tumor, whereas none of the other patients (i.e., gliomas of grade III or lower, N ¼ 9) did, demonstrated by both qualitative and quantitative assessments. The data suggest that FMISO PET may be a useful tool to distinguish glioblastomas from lower-grade gliomas. Our results, however, were slightly different from previous investigations reporting that some lower-grade gliomas (e.g., grade III) showed positive FMISO uptake. Many of these acquired the FMISO PET images 2 h after the FMISO injection, while for the study here we waited 4 h to be able to collect hypoxia-specific signals rather than perfusion signals as FMISO clearance from plasma is slow due to its lipophilic nature. No optimum uptake time for FMISO has been established, and we directly compared the 2-h vs. the 4-h images with the same patients (N ¼ 17). At 2 h, the gray matter had significantly higher standardized uptake value (SUV) than the white matter, possibly due to different degrees of perfusion but not due to hypoxia. At 4 h, there were no differences between gray and white matter without any significant increase in the noise level measured by the coefficient of variation between the 2-h and the 4-h images. At 2 h, 6/8 (75 %) of glioblastoma patients © The Author(s) 2016 Y. Kuge et al. (eds.), Perspectives on Nuclear Medicine for Molecular Diagnosis and Integrated Therapy, DOI 10.1007/978-4-431-55894-1_18 229 showed higher uptakes in the tumor than in the surrounding brain tissue, whereas at 4 h this was the case for 8/8 (100 %). In addition, at 2 h, 3/4 (75 %) of patients with lower-grade gliomas showed moderate uptakes, while at 4 h none did (0/4 or 0 %). These data indicate that 4-h images are better than 2-h images for the purpose of glioma grading. In conclusion, we evaluated the diagnostic performance of FMISO PET for gliomas and suggest that FMISO PET may be able to assist in the diagnosis of glioblastomas when PET images are acquired at 4 h post injection.