Prognostication of anaplastic astrocytoma patients: application of contrast leakage information of dynamic susceptibility contrast-enhanced MRI and dynamic contrast-enhanced MRI

Purpose To examine the applicability of contrast leakage information from dynamic susceptibility contrast-enhanced (DSC) MRI and dynamic contrast-enhanced (DCE) MRI to determine which one is the most valuable surrogate imaging biomarker for predicting disease progression in anaplastic astrocytoma (AA) patients. Materials and methods This study was approved by the institutional review board (IRB), which waived informed consent. A total of seventy-three AA patients who had undergone preoperative DCE and DSC MRI and received standard treatment, including partial resection or biopsy followed by radiation therapy, were included in this retrospective study. Based on Response Assessment in Neuro-Oncology (RANO), patients were sorted into progression (n = 21) and non-progression (n = 52) groups. Tumor boundaries were defined as high-signal intensity (SI) lesions on fluid-attenuated inversion recovery (FLAIR) imaging, where we analyzed mean pharmacokinetic parameters (K-trans, V-p, and V-e) from DCE MRI and contrast leakage information (mean extraction fraction (EF)) from DSC MRI. Results Mean V-e and mean EF were significantly higher in patients with progression-free survival (PFS) < 18 months than in those with PFS >= 18 months. For distinguishing the group with PFS < 18 months, AUC values were calculated using the mean V-e value (AUC = 0.716). The Kaplan-Meier survival analysis revealed that mean V-e value was significantly correlated with PFS. In Cox proportional-hazards regression, only the mean V-e value was found to be significantly associated with PFS. Conclusion We found that the mean V-e value based on high-SI tumor lesions on FLAIR imaging was capable of predicting outcomes of AA patients as a potential surrogate imaging biomarker.