Despite modern therapeutic advances, the median overall survival for patients with glioblastoma remains approximately 14 months.  A subset of patients progresses much more rapidly while about 1-2% of patients survive beyond 3 years from diagnosis. Yet predicting which patients are likely to progress rapidly versus slowly remains challenging.  At the transcriptomic level, database analyses have demonstrated the importance of RNA splicing variants and RNA binding protein interactions in glioblastoma pathogenesis. In this work, we explore whether mRNA expression patterns in newly diagnosed glioblastoma patients can be used to predict response to standard therapy.

Using TCGA data, we find that there are significant differences in gene expression profiles between rapid progressors and those who survive longer than 1 year.  Gene set enrichment analysis shows underexpression of markers of immune activation and cell cycle regulation, and overexpression of RAS signalling. This analysis is limited by the lack of clinical annotation available in TCGA and the small number of patients.  

These results demonstrate that mRNA expression patterns recapitulate expected markers of glioblastoma progression.  They motivate our ongoing efforts to uncover novel transcriptomic signatures in a large clinically-annotated dataset.