Prognostic value of 3'-deoxy-3'-18F-fluorothymidine ([(18)F] FLT PET) in patients with recurrent malignant gliomas.

INTRODUCTION

3'-deoxy-3'-18F-fluorothymidine ([(18)F] FLT) PET has been proven to be of value in diagnosis and assessment of glioma grading, in differentiating tumor recurrence from necrosis, in response assessment and in predicting overall survival (OS) in the primary high grade glioma. In this study, we evaluated the value of [(18)F] FLT PET-CT in predicting the OS of patients with recurrent malignant glioma.

METHODS

Fifty-six patients with recurrent malignant glioma were enrolled in this prospective study. The PET-CT and contrast-enhanced MRI scans were performed in all patients. Tumor volume was determined from both PET image (proliferative volume, PV) and MRI image (Vol-MRI). Patients were followed up clinically until death. The likelihood of using PET-derived parameters of SUVmax, tumor-to-normal (T/N) ratio, and PV to predict the OS of patients were assessed in comparison with Vol-MRI and other clinical parameters.

RESULT

The follow up periods for all patients ranged from 1.5 to 35.6 months with median of 9.8 months. Univariate analysis showed that the following parameters were significantly correlated with OS: grade of primary tumor (p=0.042), Karnofsky performance score (KPS) (p=0.041), T/N ratio (p<0.01), Vol-MRI (p=0.041), and PV (p<0.001). However, multivariate Cox regression showed that only the PV (p<0.001) and T/N ratio (p=0.001) were independent predictors. The thresholds to predict OS were 16.88 cm(3) for PV and 10.94 for T/N ratio. Kaplan-Meier analyses using these thresholds showed a significant discrimination between short and long OS groups (p<0.001).

CONCLUSION

The PV and T/N ratio of tumor on [(18)F] FLT PET-CT are independent predictors of survival in patients with recurrent malignant glioma. The PV on [(18)F] FLT PET seems to be more predictive than tumor volume on T1-weighted MRI for OS.