P03.07.A MRI signature in preoperative imaging predicts mesenchymal stem cell features and radioresistance in primary glioblastoma stem cell cultures

Abstract

Abstract Background Recently, a molecular signature describing the mesenchymal / proneural features of primary stem-cell-enriched glioblastoma cultures was identified correlating with invasiveness, radiation sensitivity and patient-outcome. However, generating and characterizing primary cultures is time-consuming and this might hamper translation into clinical concepts. The aim of this study was to evaluate the use of standard preoperative MR imaging (T1ce and T2FLAIR sequences) to predict the molecular signature and thus enable the development of clinical treatment concepts based on the molecular properties of the individual tumors. Material and Methods For 16 patients, for whom primary stem cell enriched cultures had been characterized for their mesenchymal / proneural signature and radiation sensitivity, tumor volume (hyperintense volume in T1ce), necrosis (hypointense volume inside the tumor volume) and edema (hyperintense volume in T2FLAIR) were contoured for volumetric analysis. Volume parameters were used to calculate ratios (edema/tumor and necrosis/tumor) and a MRI signature, which was then correlated with molecular parameters and patient outcome stratified by MGMT promoter methylation. Results As expected, the prognosis of patients with MGMT-promoter-methylated tumors was better (n.s.) compared to those with unmethylated MGMT promoters. Neither molecular nor imaging data were significantly different between these subgroups. In the subgroup of patients with unmethylated MGMT promoter, volumetric imaging parameters correlated with the molecular signature and cellular radiation sensitivity of the stem cell enriched cultures. This association was much weaker in the subgroup with methylated MGMT promoter. Conclusion In the subgroup of patients with unmethylated MGMT promoter, volumetric parameters on preoperative standard MR imaging might hint at the molecular properties of the respective tumor and its radiation sensitivity. This might be a clinically applicable method to stratify treatment according to molecular stem cell subtype without tissue culture and molecular analysis.