Epigenetic profiles allow for more precise predictions in brain cancer

German Cancer Research Center: Scientists from the German Cancer Research Center (DKFZ) have analyzed the DNA methylation patterns in 500 ependymomas - tumors that occur in the brain and spinal cord. They were able to distinguish nine molecular subgroups. This classification enables clinicians to better predict the widely varying courses of disease that this type of cancer can take. In addition, key molecules in the various molecular groups have been identified as promising targets for more effective drug treatment in cases where chemotherapy is virtually ineffective. Ependymomas are a group of tumors that occur in children and adults. They develop in all areas of the central nervous system (CNS): the spinal cord, the cerebellum and the cerebrum. Apart from the sites where they develop, ependymomas also vary in their biology and in the clinical courses that they take.

For classifying this type of tumor, clinicians have so far had to rely solely on histopathology, which does not provide reliable information about the course and the biology of the disease. With the goal of achieving a more useful classification of these tumors, Kristian Pajtler and colleagues in the groups of Dr. Marcel Kool and Prof. Stefan Pfister at the German Cancer Research Center (Deutsches Krebsforschungszentrum, DKFZ) have now performed a detailed molecular analysis of 500 ependymomas. They analyzed epigenetic characteristics such as DNA methylation as well as genetic alterations and gene activity.
The tumors under investigation were comprised of all histological grades and came from all regions of the brain and patients of all age groups. This allowed the researchers to perform a comprehensive classification of this heterogeneous group of cancers.
When evaluating the methylation patterns, the researchers discovered that they could categorize these tumors into nine distinct subgroups. Each of the three areas of the CNS where ependymomas are found (spinal cord, cerebellum and cerebrum) can be assigned to three of these subgroups.
In addition to their methylation profiles, the nine subgroups vary in their genetics, gene activity, patients’ age at disease onset and clinical course of disease. “The diseases are so extremely different that we almost have to assume that they arise from different cell types,” says Prof. Dr. Stefan Pfister, a pediatric oncologist and molecular biologist.
The methylation patterns stay the same during the whole course of the disease, even when it relapses. This phenomenon additionally supports classifying the cancers on this basis. Since a molecular analysis requires only minute quantities of DNA, which may even be extracted from paraffin-embedded tissue samples, this type of test would lend itself to being used in clinical practice.
Molecular analysis reflects the various clinical courses of this disease more precisely than classification based on microscopic tissue analysis does. “The fact that patients with an extremely poor prognosis, mostly children, can be assigned primarily to two of the nine groups is a very good indicator that our molecular classification is clinically relevant,” says study head Marcel Kool.
When the researchers analyzed gene activity, they additionally discovered that varying signaling pathways are overactive in the tumor cells of the various groups. Targeted agents directed against several key proteins in these signaling pathways could thus be used to slow down tumor growth. “Particularly in the two tumor groups for which we are still in need of more effective treatment options, the analysis of DNA methylation may be the key to new therapies,” says Stefan Pfister.