CBMT-15. METABOLIC AND TRANSCRIPTIONAL PROFILES OF GBM INVASION: COMPARISON OF PATIENTS AND PAIRED PATIENT DERIVED XENOGRAFTS USING1H MAGNETIC RESONANCE SPECTROSCOPY AND IMAGING (7T AND 14T) AND RNA-SEQUENCING. (5th November 2018)

• Record Type: Journal Article
• Authors: Hegi, Monika; Cudalbu, Cristina; Bady, Pierre; Lai, Marta; Xin, Lijing; Gusyatiner, Olga; Hamou, Marie-France; Lepore, Mario; Daniel, Roy; Gruetter, Rolf; Hottinger, Andreas
• Is Part Of: Neuro-oncology. Volume 20(2018)Supplement 6
• Page Start: vi35
• Page End: vi35
• Publication Date: 2018-11-05
• Subjects: Brain Neoplasms -- Periodicals; Brain -- Tumors -- Periodicals; Brain -- Cancer -- Periodicals; Nervous system -- Cancer -- Periodicals; 616.99481
• DOI: 10.1093/neuonc/noy148.134
• Languages: English
• ISSNs: 1522-8517

Abstract

Abstract: BACKGROUND: Glioblastoma (GBM) are notorious for their invasive behavior. Little is known about the biology of the tumor cells and the microenvironment at the invasive front that is highly relevant for recurrence. These cells are technically difficult to visualize, lie usually outside the resected and irradiated area, and are behind the blood brain barrier that renders them difficult to treat. We present patient specific metabolic and transcriptomic features of invasive growth using corresponding patient derived xenografts (PDX). METHODS: Patients with suspected GBM were enrolled and underwent 1 H magnetic resonance spectroscopy and imaging ( 1 H-MRS/I) at 7 Tesla. Tissue obtained at the subsequent resection was dissociated and transplanted orthotopically into mice (n=4 to 6). Mice were followed longitudinally by 1 H-MRS/I (14T). The PDX, the corresponding tissue of the contralateral side, and the original tumors were subjected to RNA sequencing. Tumor versus host (mouse) derived sequencing reads are computationally separated. RESULTS: Diffuse xenografts developed for 8 (IDHwt) of 9 patients within 2–5 months. 1 H-MRS at ultrahigh fields allowed reliable quantification of 22 metabolites. The temporal changes of the metabolite signatures characterized the kinetics of invasive growth on both, the injected and the contralateral side. At end stage the signatures corresponded well with histological findings. Migration to the contralateral side ranged in a patient dependent manner between < 1 to > 50% 50% of tumor cells. Comparison of MRS derived metabolite signatures at end stage and the corresponding human signatures compared best with voxels measured outside the core of the human tumor. Associations between the metabolite signatures and the transcriptome in the xenografts and the host will be discussed. CONCLUSION: Integration of metabolic profiles and gene expression of the tumor and the invaded brain may provide insights and tools for in vivo monitoring of treatment response in the infiltration zone. GRANTS: Swiss Bridge Award, Swiss Cancer Ligue