Speaker: Marcel Kool
Department: Pediatric neuro-oncology
Subject: Molecular classification of pediatric brain tumours reveals new entities.
Location: Erasmus MC Room Ae-406
Author: Kristian Blom
Brain tumours, one of the worst types of cancer. In 2007 the WHO (World Health Organization) came with a classification of brain tumours on histological aspects and grades, called the World Health Organization Classification of Tumours. These classifications are important for clinical treatment of brain tumours, since every brain tumour needs a different treatment. Unfortunately, just by looking at brain tumour tissue under the microscope isn’t enough for proper classification, we also have to look to the brain tumour tissue at molecular level. Marcel Kool, the speaker of this seminar, is one of the researchers of the German Cancer Research Center (DKFZ: Deutsches Krebsforschungzentrum). In his research group he classifies pediatric brain tumours – tumours that occur in children – at the molecular level.
There are different ways to identify molecular subgroups (mRNA expression, immunohisto chemistry), but the most used and successive one is looking at the DNA methylation profile. DNA methylation is a cellular process whereby methyl groups are binding to the DNA. This binding alters the structure of the DNA itself (not the DNA sequence), whereby the function of the DNA gets modified. DNA methylation is very important in gene regulation; the process where external factors switches a gene on/off. When a mutation in DNA methylation occurs, it is possible that a cell will become a tumour cell.
An example of pediatric brain tumours classification are the ependymal tumours. These kind of tumours occur in different regions of the central nervous system at different ages. Marc and his research group collected 500 ependymal tumour samples, classified them by looking at the DNA methylation profile, and they concluded that there are 9 subgroups (see image 1-A) in the ependymal tumours.
A remarkable observation was that at the molecular level there wasn’t any distinction between the DNA methylation profile of grade II and grade III (more aggressive) ependymal tumours. This intended that based on molecular distinction, the clinical treatment of both grade II and grade III ependymal tumours should be the same. Of course this assumption must be verified by doing more research, because not everything can be said by only observing the molecular nature of a tumour.
Another important discovery of the ependymal tumour is that when it reoccurs, it will remain in the same molecular subgroup as first. For clinical aspects this is very useful, since unfortunately relapses in brain tumours occur a lot.
Last, Marc and his group found out that there are two highly aggressive subgroups of the ependymal tumour: PF-EPN-A and ST-EPN-RELA (see image 1-B). Patients with one of these tumours should get the maximum treatment. Remarkable is that both have very distinct methylation profiles, which suggest that different mutations can cause highly aggressive ependymal tumours.
The ideal view for the future is that every patient with a brain tumour can be healed without relapses. But to reach this goal, we have to know the classification of a tumour to give every patient an optimized treatment. That’s why the research of Marc Kool is very important, since they are making an algorithm for brain tumours classification. Imagine in the future that neuroscientist make a methylation profile of a patient’s tumour, upload it onto a website/app, and in just a few minutes they receive information about the kind of tumour they are dealing with. This will definitely increase the success rate of brain tumour treatment.
Since a DNA methylation profile only express one part of the molecular nature of a tumour, I think it is really important for future research to combine different aspects of the molecular nature for more accurate tumour classification. Think about physical features like heat, forces, pressure etc. These factors could vary for different tumours. The more differences we can find in different aspects of a tumour cell (histological, physical, biological), the more accurate the tumour classification would be.