Validation of neuroradiologic response assessment in gliomas: Measurement by RECIST, two-dimensional, computer-assisted tumor area, and computer-assisted tumor volume methods
Image Interpretation, Computer-Assisted
Significant limitations are associated with the use of standard radiographic measurements as indicators of response in glioma therapy trials. The Response Evaluation Criteria in Solid Tumors (RECIST) were recently introduced in an attempt to standardize and simplify assessment of response to treatment in cancer clinical trials. However, their applicability in gliomas has been assessed in only a very small number of patients. Our aim was to validate radiographic response assessment in newly diagnosed glioma patients. Sixty-seven newly diagnosed glioma patients participating in nine North Central Cancer Treatment Group glioma trials were included; 565 MRI scans were analyzed. All scans were performed with the same technique. Kappa statistics were calculated to determine agreement between assessment methods. Cox proportional hazards analyses and time-dependent Cox models were used to assess the association between different measurement methods and overall survival. Results showed agreement between the one-dimensional (1D) and two-dimensional (2D) measurements both for T2 images and for gadolinium-enhanced images. Comparison of duration of response and time to progression as assessed by eight different methods showed similarity in response assessments by 1D, 2D, area, and volume gadolinium measurements. In contrast, time to progression was significantly shorter when assessed by 1D-T2 or 2D-T2 images as compared to area-T2 or volume-T2 images. This set of data indicates that RECIST could be used instead of 2D imaging for response assessment in newly diagnosed glioma trials. Overall, responses as determined by any tumor measurement method did not correlate with patient survival for either enhancing or nonenhancing tumors, although the small number of responders limits definitive conclusions. Time-dependent Cox models demonstrated that, in contrast to the case of nonenhancing tumors, progression as determined by 1D, 2D, area, and volume measurements in gadolinium-enhanced images was predictive of survival of patients with enhancing tumors.