Research focuses on the identification of clinically relevant molecular subgroups and functional characterization of genes involved in glioma initiation and/or progression using state of the art high-throughput genomic approaches.

Identification of clinically relevant molecular subgroups
Correct classification of gliomas is important because it directly affects the path of treatment to be followed. However, classification of gliomas based on histological appearance is difficult and subject to significant inter-observer variability. This variability can result in the assigning of a prognostically favorable lower-grade glioma into a poor prognostic cohort (“false-positive GBM”), and in the assigning of a prognostically less favorable higher-grade glioma into a good prognostic group (“false-negative GBM”).

To improve current classification standards, we have performed gene expression profiling on 273 glioma samples of all histological subtypes and grades. We identified seven distinct molecular subgroups “intrinsic molecular subtypes” of glioma (IGS) that are different from histological subgroups and correlate better to patient survival. The power of intrinsic subtyping is demonstrated by its ability to identify a subset of prognostically favorable tumors within an external dataset that contains only histologically confirmed glioblastomas. Conversely, molecular clustering was also able to distinguish grade II gliomas with a prognosis comparable to GBMs (figure 1). Our data provides compelling evidence that expression profiling is a more accurate and objective method to classify gliomas than histological classification. The prognostic value of these intrinsic glioma subtypes was recently validated on randomized clinical trial material. Molecular classification therefore may aid diagnosis and can guide clinical decision making.
Selection of patients who benefit most from particular therapeutic regimens helps to improve treatment efficacy and to potentially avoid toxicity in patients who are unlikely to benefit from that therapy anyway. Since several projects make use of material from randomized clinical trials, this material allows the identification of markers that predict response to treatment. Because intrinsic glioma subtypes represent molecularly distinct entities of glioma, it is possible that each subtype requires its own treatment regimen. Indeed, our results demonstrate that tumors assigned to IGS-9 show benefit form adjuvant PCV chemotherapy when added to radiotherapy; other molecular subtypes do not show this benefit. We are currently evaluating a predictive role for intrinsic subtypes in the EORTC 22033 and Belob clinical trials.

An 85 gene signature has been develloped that can be used to classify samples to one of the seven intrinsic subtypes. The signature genes can be found here.

Figure 1: Molecular profiling identifies prognostically favorable gliomas within an independent group of GBMs (left) and prognostically poor gliomas within a group of grade II gliomas (right). All gliomas were separated by their intrinsic gene expression profiling into molecular subgroups that correlate with survival.

Functional analysis of genes involved in glioma initiation and/or progression
Genetic changes involved in glioma-genesis or progression are potential therapeutic targets and a number of such changes have been identified in gliomas. In GBMs, causal genes include the oncogenes EGFR, MDM2, and CDK4 and the tumor suppressor genes PTEN, CDKN2A and NF1. In lower grade gliomas, causal cancer genes include IDH1, TP53, CIC, FUBP1, ATRX and NOTCH1

We are currently employing next generation sequencing (both DNA and RNA-seq) in combination with other state of the art 'omics' techniques (including high-content, high-throughput confocal imaging, mass spectrometry and reverse phase protein arrays) to functionally characterize the causal genes of the various subtypes of gliomas. At present, the functional analysis is focussed on mutations in the IDH1 and EGFR genes. Our aim is to determine the role of these proteins in glioma initiation and/or progression, explain why specific mutations accumulate in distinct tumor subtypes and identify novel drugs that target these genetic changes. For example, the epidermal growth factor receptor (EGFR) gene is a key oncogene that is mutated in many tumors including lung adenocarcinomas (LUAD) and glioblastomas (GBM). In LUAD, around 90% of all EGFR mutations comprise of either short in-frame deletions in exon 19 (in particular around residues 747-750) or the L858R missense mutation in exon 21. In GBMs, the EGFR gene is also mutated, but here the gene is high copy number amplified with levels sometimes exceeding 100 copies per cell. This initial amplification is then followed by the acquisition of additional mutations that including intragenic deletions, point mutations and gene-fusions.


Figure 2: Distribution of EGFR mutations in glioblastomas and lung cancer. As can be seen, each tumor-type has its own distribution of mutations in EGFR with those in GBMs concentrating on the C-terminal extracellular domain. EGFRvIII is the most common mutation in GBMs, present in ~50% of tumors, and is an in-frame deletion of exons 2-7.

  1. Gravendeel et al. Cancer Res. (2009), 69:9065-72.
  2. Erdem-Eraslan et al. J. Clin Oncol (2013), 31: 328-36 
  3. Erdem-Eraslan et al. Eur J Cancer (2015), 51: 893-903
  4. Gao et al, Cancers (2018), 10:E489