ADC quantification to classify patients candidate to receive bevacizumab treatment for recurrent glioblastoma

Abstract

Background Recurrent high-grade gliomas progressing after surgery and temozolomide plus radiation therapy have traditionally been treated using antiangiogenic drugs as the first-line therapy. Since the phase 3 EORTC 26101 trial showed no significant benefit of administering antiangiogenic drugs, the need to identify a biomarker to classify subgroups of potential responders has increased. Purpose To investigate the feasibility of using apparent diffusion coefficient as a predictor of the response of recurrent high-grade gliomas to bevacizumab or classifier for patients showing better response. Material and Methods This retrospective study analyzed magnetic resonance images obtained from 39 patients at the time of high-grade glioma progression who were treated using bevacizumab. The apparent diffusion coefficient maps were quantified and modelled as a mixture of Gaussian functions. The correlation of their descriptors and the time to second progression was studied. Log-rank tests were performed to determine the power of these descriptors as the classifiers for patients exhibiting better survival. Results None of the descriptors showed correlation with time to second progression (r < 0.35) but several of them stratified subgroups showing a better time to second progression passing log-rank tests ( P < 0.02). Conclusion Apparent diffusion coefficient cannot be used to predict the time to second progression of recurrent high-grade gliomas treated with bevacizumab, but it can stratify groups with better time to second progression distributions.

Machine-Learning-Based Radiomics MRI Model for Survival Prediction of Recurrent Glioblastomas Treated with Bevacizumab

“…Moreover, the correlation threshold in this analysis suggests that these clinical variables offer complementary data to the radiomics, which may lead to the elaboration of other advanced models using different sources of data. Various studies have shown that the apparent diffusion coefficient (ADC) from diffusion MR could represent a good predictor of survival in GBM patients treated with bevacizumab [ 36 , 37 , 38 , 39 ]. In addition, perfusion-derived biomarkers were explored [ 40 , 41 , 42 ], but it should be mentioned that these methods depend largely on the type of the MRI machine, the sequence processing software, and the calculation algorithms and, therefore, are more difficult to apply to different centres in routine clinical practice.…”