Early assessment of recurrent glioblastoma response to bevacizumab treatment by diffusional kurtosis imaging: a preliminary report

Purpose The purpose of this preliminary study is to apply diffusional kurtosis imaging to assess the early response of recurrent glioblastoma to bevacizumab treatment. Methods This prospective cohort study included 10 patients who had been diagnosed with recurrent glioblastoma and scheduled to receive bevacizumab treatment. Diffusional kurtosis images were obtained from all the patients 0-7 days before (pre-bevacizumab) and 28 days after (post-bevacizumab) initiating bevacizumab treatment. The mean, 10th, and 90th percentile values were derived from the histogram of diffusional kurtosis imaging metrics in enhancing and non-enhancing lesions, selected on post-contrast T1-weighted and fluid-attenuated inversion recovery images. Correlations of imaging measures with progression-free survival and overall survival were evaluated using Spearman's rank correlation coefficient. The significance level was set at P < 0.05. Results Higher pre-bevacizumab non-enhancing lesion volume was correlated with poor overall survival (r = -0.65, P = 0.049). Higher post-bevacizumab mean diffusivity and axial diffusivity (D-parallel to, D-parallel to 10% and D-parallel to 90%) in non-enhancing lesions were correlated with poor progression-free survival (r = -0.73, -0.83, -0.71 and -0.85; P < 0.05). Lower post-bevacizumab axial kurtosis (K-parallel to 10%) in non-enhancing lesions was correlated with poor progression-free survival (r = 0.81, P = 0.008). Conclusions This preliminary study demonstrates that diffusional kurtosis imaging metrics allow the detection of tissue changes 28 days after initiating bevacizumab treatment and that they may provide information about tumor progression.