Pre-operative MRI radiomics model non-invasively predicts key genomic markers and survival in glioblastoma patients.

PURPOSE: Although glioblastoma (GBM) is the most common primary brain malignancy, few tools exist to pre-operatively risk-stratify patients by overall survival (OS) or common genetic alterations. We developed an MRI-based radiomics model to identify patients with EGFR amplification, MGMT methylation, GBM subtype, and OS greater than 12 months. METHODS: We retrospectively identified 235 patients with pathologically confirmed GBMs from the Cancer Genome Atlas (88; TCGA) and MD Anderson Cancer Center (147; MDACC). After two neuroradiologists segmented MRI tumor volumes, we extracted first-order and second-order radiomic features (gray-level co-occurrence matrices). We used the Maximum Relevance Minimum Redundancy technique to identify the 100 most relevant features and validated models using leave-one-out-cross-validation and validation on external datasets (i.e., TCGA). Our results were reported as the area under the curve (AUC). RESULTS: The MDACC patient cohort had significantly higher OS (22 months) than the TCGA dataset (14 months). On both LOOCV and external validation, our radiomics models were able to identify EGFR amplification (all AUCs > 0.83), MGMT methylation (all AUCs > 0.85), GBM subtype (all AUCs > 0.92), and OS (AUC > 0.91 on LOOCV and 0.71 for TCGA validation). CONCLUSIONS: Our robust radiomics pipeline has the potential to pre-operatively discriminate common genetic alterations and identify patients with favorable survival.

Src homology domain-containing phosphatase 2 suppresses cellular senescence in glioblastoma.

BACKGROUND: Epidermal growth factor receptor (EGFR) signalling is frequently altered during glioblastoma de novo pathogenesis. An important downstream modulator of this signal cascade is SHP2 (Src homology domain-containing phosphatase 2). METHODS: We examined the The Cancer Genome Atlas (TCGA) database for SHP2 mutations. We also examined the expression of a further 191 phosphatases in the TCGA database and used principal component and comparative marker analysis available from the Broad Institute to recapitulate the TCGA-defined subgroups and identify the specific phosphatases defining each subgroup. We identified five siRNAs from two independent commercial sources that were reported by the vendor to be pre-optimised in their specificity of SHP2 silencing. The specificity and physiological effects of these siRNAs were tested using an in vitro glioma model. RESULTS: TCGA data demonstrate SHP2 to be mutated in 2% of the glioblastoma multiforme's studied. Both mutations identified in this study are likely to be activating mutations. We found that the four subgroups of GBM as defined by TCGA differ significantly with regard to the expression level of specific phosphatases as revealed by comparative marker analysis. Surprisingly, the four subgroups can be defined solely on the basis of phosphatase expression level by principal component analysis. This result suggests that critical phosphatases are responsible for the modulation of specific molecular pathways within each subgroup. Src homology domain-containing phosphatase 2 constitutes one of the 12 phosphatases that define the classical subgroup. We confirmed the biological significance by siRNA knockdown of SHP2. All five siRNAs tested reduced SHP2 expression by 70-100% and reduced glioblastoma cell line growth by up to 80%. Profiling the established molecular targets of SHP2 (ERK1/2 and STAT3) confirmed specificity of these siRNAs. The loss of cell viability induced by SHP2 silencing could not be explained by a significant increase in apoptosis alone as demonstrated by terminal deoxyribonucleotidyl transferase-mediated nick-end labelling and propidium iodide staining. Src homology domain-containing phosphatase 2 silencing, however, did induce an increase in ß-galactosidase staining. Propidium iodide staining also showed that SHP2 silencing increases the population of glioblastoma cells in the G1 phase of the cell cycle and reduces the population of such cells in the G2/M- and S-phase. CONCLUSION: Src homology domain-containing phosphatase 2 promotes the growth of glioblastoma cells by suppression of cellular senescence, a phenomenon not described previously. Selective inhibitors of SHP2 are commercially available and may be considered as a strategy for glioblastoma therapy.