P06.11 Classification of high-grade glioma using Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry imaging (MALDI MSI): interim results of the GLIOMIC study

Abstract

Matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI MSI) allows to determine spatial localization of molecules along tissue sections. The aim of the Gliomic study is to classify high-grade gliomas (HGG) by MALDI-MSI according to their prognosis. Until now, samples from 13 patients have been analyzed after tumor resection. According to the OMS 2007 classification, 3 tumors were grade 3 (2 oligoastrocytoma and 1 astrocytoma) and 10 were grade 4. An isocitrate dehydrogenase 1 (IDH1) mutation was found in 31% of the cases, a O(6)-methylguanine-DNA methyltransferase (MGMT) promoter methylation in 20% (median value: 5%, range 2-93%). Frozen tissue sections of 12 mm thickness were realized and matrix was then deposited homogeneously with a manual sprayer to cover the entire surface. Tissues were prepared foranalysis of low-weighted molecules, particularly lipids. Tumor-specific mass spectra were acquired with 70 mm spatial resolution. After molecular image acquisition, matrix was removed, tissues were stained and an notated by a pathologist. The imaging datasets were spatially segmented by clustering all spectra by their molecular profiles similarity. Individual tissue segmentations were compared with histological annotations and revealed new regions of interest. A low concordance between the histological annotations and the MS imaging data was found for tumor regions. Global segmentations were generated using all the tissues and revealed 3molecularly different groups. Known low infiltrated regions and necrotic regions were still grouped together respectively after this global segmentation. Pre-necrotic regions were divided between necrotic and tumor regions. Moreover, we highlighted sub- regions within the tumor are a which were not histologically an notated. This analysis provided new insights into HGG tissue organization. The data obtained from MALDIMSI analysis may provides pecific information and allow amore accurate diagnosis. A total of 21patients have been enrolled and updated results will be presented taking account histologicalandmolecular biologydata.