Team of scientists uncovers genetic mutations linked to aggressive brain tumor

Scientists at Washington University School of Medicine in St. Louis, working as part of a large-scale federally funded research collaboration, have discovered new genetic mutations and molecular pathways underlying glioblastoma, the most common form of brain cancer and the most aggressive.

The first results of this comprehensive genetic study are published Sept. 4 in the advance online edition of the journal Nature. The findings lay the foundation for developing new ways to diagnose and treat the deadly disease.

“This is extremely exciting,” says Richard K. Wilson, Ph.D., director of the Genome Sequencing Center at the School of Medicine. “This is exactly why we sequenced the human genome – so that we would have a reference point for looking at individual human genomes from people who have a disease like cancer to try to understand what’s gone wrong. Only then can we start to think about how to use that information to better diagnose their disease, delineate their cancer subtype and identify more targeted therapies.

The researchers identified numerous genetic mutations involved in glioblastoma, including three previously unrecognized mutations that occur frequently, and defined core molecular pathways that are disrupted. Among the most exciting results is an unexpected observation that points to a potential mechanism of resistance to a common chemotherapy drug used for brain cancer.

More than 21,000 new cases of brain cancer are predicted in the United States this year, with more than 13,000 people likely to die from the disease. Glioblastoma, the type of brain cancer most often found in adults, is fast-growing. Most patients with glioblastoma die of the disease within 14 months of diagnosis.

The scientists, all part of The Cancer Genome Atlas (TCGA) Research Network, analyzed the complete sets of DNA, or genomes, of tumor samples donated by 206 patients with glioblastoma. The work complements and expands upon a parallel study by Johns Hopkins researchers of 22 glioblastoma tumors, which was also published Sept. 4 in the journal Science.

Genome scientists at the School of Medicine brought their experience in tumor sequencing to the glioblastoma project by formulating the sequencing strategy, contributing tumor sequence data and then developing the computer software and tools to analyze the resulting genetic data.

Wilson likened finding the genetic changes involved in glioblastoma to looking for a needle in a haystack. “All humans have considerable DNA variations at the level of the genome, and that is completely normal,” Wilson explained. “We had to find the few DNA base changes – the needles – linked to glioblastoma in a haystack of six billion bases of DNA that makeup the full complement of genetic material in humans.”

Like most cancers, glioblastoma arises from changes that accumulate in cells’ DNA over the course of a person’s life – changes that may eventually lead to the cells’ uncontrolled growth. However, until recently, scientists have understood little about the precise nature of these DNA changes and their impact on key biological pathways that are important to the development of new interventions.

“These impressive results from TCGA provide the most comprehensive view to date of the complicated genomic landscape of this deadly cancer. The more we learn about the molecular basis of glioblastoma, the more swiftly we can develop better ways of helping patients with this terrible disease,” said Elias A. Zerhouni, M.D., director of the National Institutes of Health (NIH). “Clearly, it is time to move ahead and apply the power of large-scale, genomic research to many other types of cancer.”

The National Cancer Institute and the National Human Genome Research Institute, both part of the NIH, initiated TCGA in 2006 to unlock the genetic secrets underlying cancer by using the powerful tools of DNA sequencing. Washington University School of Medicine is one of three large-scale genome sequencing centers involved in TCGA, which also is charged with identifying the genetic changes underlying lung and ovarian cancer.

In the Nature paper, the TCGA Research Network describes the interim results of its analyses of glioblastoma, the first type of cancer to be studied as part of the research collaboration. The pioneering work pulled together and integrated multiple types of data generated by several genome characterization technologies from investigators at 18 different participating institutions and organizations. The data include small changes in DNA sequence, known as genetic mutations; larger-scale changes in chromosomes, known as copy number variations and chromosomal translocations; the levels of protein-coding RNA being produced by genes, known as gene expression; patterns of how certain molecules, such as methyl groups, interact with DNA, known as epigenomics; and information related to patients’ clinical treatment.

TCGA researchers sequenced 601 genes in glioblastoma samples and matched control tissue, uncovering three significant genetic mutations not previously reported to be common in glioblastoma. The affected genes were: NF1, a gene previously identified as the cause of neurofibromatosis 1, a rare, inherited disorder characterized by uncontrolled tissue growth along nerves; ERBB2, a gene that is well-known for its involvement in breast cancer; and PIK3R1, a gene that influences activity of an enzyme called PI3 kinase that is deregulated in many cancers. PI3 kinase already is a major target for therapeutic development. The discovery of frequent mutations in the PIK3R1 gene means that glioblastoma patients’ responses to PI3 kinase inhibitors may be dictated by whether or not their tumors have mutated versions of the gene.

The TCGA team combined sequencing data with other types of genome characterization information, such as gene expression and DNA methylation patterns, to generate an unprecedented overview that delineated core biological pathways potentially involved in glioblastoma. The three pathways, each of which was found to be disrupted in more than three-quarters of glioblastoma tumors, were: the CDK/cyclin/CDK inhibitor/RB pathway, which is involved in the regulation of cell division; the p53 pathway, which is involved in response to DNA damage and cell death; and the RTK/RAS/PI3K pathway, which is involved in the regulation of growth factor signals.

The pathway mapping promises to be particularly informative for researchers working to develop therapeutic strategies that are aimed more precisely at specific cancers or that are better tailored to each patient’s particular subtype of tumor.

For example, a patient whose tumor has genetic alterations at one point in the CDK pathway might benefit from a drug that blocks CDK, while patients with mutations at another point in the same pathway might be predicted not to respond to such drugs. Similarly, while some drugs already used for glioblastoma target the RTK pathway, the new findings suggest a need to tailor therapeutic cocktails to particular patterns of mutations in genes involved in that pathway.

The three pathways were interconnected and coordinately deregulated in most of the glioblasoma tumors analyzed. Therefore, combination therapies directed against all three pathways may offer an effective strategy, the researchers noted.

One particularly exciting finding with the potential for rapid clinical impact centers on the MGMT gene. Physicians already know patients with glioblastoma tumors that have an inactivated, or methylated, MGMT gene respond better to temozolomide, an alkylating chemotherapy drug commonly used to treat glioblastoma. By integrating methylation and sequencing data with clinical information about sample donors, TCGA’s multi-dimensional analysis found that in patients with MGMT methylation, alkylating therapy may lead to mutations in genes that are essential for DNA repair, commonly known as mismatch repair genes. Such mutations then lead to the subsequent emergence of recurrent tumors that contain an unusually high number of DNA mutations, and that may be resistant to chemotherapy treatment. If follow-up studies confirm such a mechanism, researchers say first- or second-line treatments for such glioblastoma patients may involve therapies designed to target the results of combined loss of MGMT and mismatch-repair deficiency. The new findings also may help clinical researchers figure out the best ways to combine alkylating chemotherapy drugs with the next generation of targeted therapeutics.

“This represents another major step towards our ultimate goal of using information about the human genome to improve human health,” said NHGRI Acting Director Alan E. Guttmacher, M.D. “It’s thrilling to see what the cancer and genomics research communities can achieve through working together in a collaborative manner. I am confident that this paper is just the first of many exciting results that TCGA will generate.”

As in the Human Genome Project, TCGA data are being made rapidly available to the research community through a database, http://cancergenome.nih.gov/dataportal. The database provides access to public datasets, and with required review and approval, allows researchers access to more in-depth data.

The TCGA Research Network members who chaired the large TCGA committee that wrote the Nature paper were Lynda Chin, M.D., and Matthew Meyerson, M.D., Ph.D., of Dana-Farber Cancer Institute and Harvard Medical School. Dr. Meyerson is also affiliated with the Broad Institute of MIT and Harvard.


Washington University School of Medicine’s 2,100 employed and volunteer faculty physicians also are the medical staff of Barnes-Jewish and St. Louis Children’s hospitals. The School of Medicine is one of the leading medical research, teaching and patient care institutions in the nation, currently ranked third in the nation by U.S. News & World Report. Through its affiliations with Barnes-Jewish and St. Louis Children’s hospitals, the School of Medicine is linked to BJC HealthCare.