Correlation between 18F-fluoromisonidazole PET and expression of HIF-1α and VEGF in newly diagnosed and recurrent malignant gliomas

Purpose Hypoxia and its consequences at the molecular level promote tumour progression and affect patient prognosis. One of the main early cellular events evoked by hypoxia is induction of hypoxia-inducible factor 1 (HIF-1) and subsequent upregulation of vascular endothelial growth factor (VEGF). In this study we sought to determine whether hypoxia detected by F-18-fluoromisonidazole (FMISO) PET accurately reflects the expression of HIF-1 alpha and VEGF in the tumour and can be used as a biomarker of antiangiogenic treatment and as a prognostic factor in newly diagnosed and recurrent malignant gliomas. Methods Enrolled in this study were 32 patients with newly diagnosed glioma and 16 with recurrent glioma of grade III or grade IV. All the patients had undergone FMISO PET preoperatively. The maximum tumour-to-blood FMISO activity ratio (T/B-max) was used to evaluate the degree of tumour hypoxia and the hypoxic volume (HV) was calculated using a tumour-to-blood FMISO uptake ratio of >= 1.2. Immunohistochemical expressions of HIF-1 alpha and VEGF were evaluated semiquantitatively using the immunoreactivity score (IRS, scores 0 to 12) and the correlation was examined between IRS of HIF-1 alpha or VEGF and FMISO uptake of the tumour (SUVtumour) using navigation-based sampling. Survival was estimated using the Kaplan-Meier method in relation to the T/B-max and the HV. Results The T/B-max and the HV in grade IV gliomas were significantly higher than in grade III gliomas (P < 0.01 and P < 0.01, respectively). Moderate to strong HIF-1 alpha and VEGF expression was observed in the majority of malignant gliomas. The IRS of HIF-1 alpha and VEGF in the tumour were not significantly different between grade III and grade IV gliomas. The IRS of HIF-1 alpha in the tumour did not correlate with the SUVtumour of FMISO in either newly diagnosed or recurrent glioma. There was a significant but weak correlation between the IRS of VEGF and the SUVtumour of FMISO in newly diagnosed glioma, but not in recurrent glioma. The overall survival time in patients with a small HV and a low FMISO T/B-max was significantly longer than in those with a large HV and a high FMISO T/B-max (P < 0.01 and P < 0.05, respectively). Conclusion Preoperative FMISO uptake is significantly correlated with the expression of VEGF in the tumour and might be used as a biomarker of antiangiogenic treatment in newly diagnosed malignant gliomas. However, caution is required because the correlation was weak and there was a large overlap of FMISO uptake between glioma with high and low VEGF expression. In addition, hypoxia determined by FMISO PET appears to be a suitable biomarker for predicting a highly malignant tumour and a poor prognosis in patients with malignant glioma.