Potential (18)F-RGD PET/CT and DCE-MRI Imaging-Based Biomarkers for Postoperative Survival Prediction Among Patients With Newly Diagnosed Glioblastoma Treated With Bevacizumab and Chemoradiotherapy

PURPOSE: To investigate the ability of potential imaging biomarkers based on (18)F-AlF-NOTA-PRGD2 positron emission tomography/computed tomography ((18)F-RGD PET/CT) and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) imaging to predict the response to bevacizumab combined with conventional therapy in postoperative newly diagnosed glioblastoma. METHODS: Twenty patients with newly diagnosed with glioblastoma after surgery were prospectively enrolled to receive bevacizumab plus conventional concurrent radiotherapy and temozolomide (CCRT). (18)F-RGD PET/CT and DCE-MRI were performed at baseline, week 3, and week 10 for each patient. Statistical methods included the analysis of variance (ANOVA), Kaplan–Meier method and Cox proportional hazard analysis. RESULTS: All patients completed CCRT plus bevacizumab therapy without interruption. The median follow-up time was 33.9 months (95% confidence interval [CI], 28.3-39.5 months). The median progression-free survival (PFS) and overall survival (OS) was 9.66 months (95% CI, 6.20-13.12 months) and 15.89 months (95% CI, 13.89-17.78), respectively. Treatment was generally well tolerated, and there were no Treatment emergent adverse events (TEAEs) with a toxicity grade equal to or exceeding 3 or that led to termination of treatment or patient death.Over the treatment interval of bevacizumab therapy from week 3 to week 10, patients with a large decrease of SUVmean was associated with a better PFS with a hazard ratio (HR) of 6.562, 95% CI (1.318-32.667), p=0.022. According to Kaplan-Meier analysis, patients with a decrease in the SUVmean of more than 0.115 on (18)F-RGD PET/CT had a longer PFS than those with a decrease in the SUVmean of 0.115 or less (12.25 months vs.7.46 months, p=0.009). For OS, only a small decrease of Ktrans was also found to have certain prognostic value (HR=0.986, 95% CI (0.975-0.998), p=0.023). Patients with a decrease in Ktrans larger than 37.03 (min-1) on DCE-MRI had worse OS than those with a decrease in Ktrans of 37.03 (min-1) or less (15.93 months vs. 26.42 months, p=0.044). CONCLUSION: (18)F-RGD PET/CT and DCE-MRI may be valuable in evaluating the response of glioblastoma to treatment with the combination of bevacizumab and CCRT, with a greater decrease in SUV(mean) predicting better PFS as well as a small decrease in K(trans) predicting improved OS.