UCSF scientists have discovered commonly inherited genetic risks for the deadliest types of brain tumors, marking the first time such a link has been definitively established. At least 20 percent of these advanced brain cancers may be attributable to the newly discovered genetic risks, according to UCSF scientists led by genetic epidemiologists Margaret Wrensch, PhD, Jeffrey Chang, MD, PhD, and John Wiencke, PhD. Their study appears in the July 5 online edition of the scientific journal Nature Genetics. “These results clearly indicate that genetic susceptibility is an important risk factor for this disease,” Wrensch says. Wiencke, Wrensch and Chang, a postdoctoral fellow, studied high-grade brain tumors known as gliomas, more specifically glioblastomas and anaplastic astrocytomas. These forms account for about three-quarters of all malignant brain tumors.
Wrensch has searched for brain tumor risks for more than 20 years, heading up the San Francisco Adult Glioma Study and, more recently, leading a UCSF project in a multidisciplinary brain tumor research initiative funded by the National Cancer Institute (NCI). “Now that we have identified these solid risk factors, it opens up new areas of research,” she says. Follow-up research is expected to shed light on biological mechanisms of brain cancer risk, about which relatively little is known. The only previously confirmed risks for the disease include exposure to high-dose radiation or rare familial syndromes, neither of which accounts for many cases of the disease, Wrensch notes.
Wrensch, Wiencke and Chang identified one high-risk genetic variant and one low-risk variant on chromosome 20, and another high-risk variant on chromosome 9. As part of the study, all three of these findings were confirmed by Robert Jenkins, MD, PhD, and Ping Yang, MD, PhD, working with data and samples from a separate group of patients and controls at the Mayo Clinic in Rochester, Minnesota. A third high-risk variant on chromosome 5 did not quite reach statistical significance in the confirmatory study. However, all three of these regions were independently identified in a separate study published in the same edition of Nature Genetics by a research team from the University of Texas M.D. Anderson Cancer Center in Houston and the Institute of Cancer Research in England.
Inheriting a copy of one of the high-risk or low-risk genetic variants appears to increase or decrease risk for glioma by roughly 50 percent, according to the UCSF and Mayo Clinic results. Brain cancers are not common, although they account for a disproportionate number of years of life lost due to cancer. Cancers of the brain and central nervous system account for about 18,000 cancer cases and 13,000 deaths in the United States each year. Although the genetic variants with newly identified risk are common – each is present in at least 10 percent of the US population – the vast majority of people who inherit one or even two copies of these higher-risk variations will never be stricken with a debilitating brain tumor.
While it is not practical, therefore, to screen the population to determine who has the high-risk variants, the findings might prove more important for individuals who have been diagnosed with brain tumors. Wrensch, Wiencke and their clinical colleagues are interested in tracking whether these genetic variants correspond to clinical outcome, and whether they are associated with the effectiveness of experimental treatments for gliomas. New treatments are sorely needed. A majority of brain tumors are advanced, grade 4 tumors by the time of diagnosis. Many common symptoms are not specific to the disease. An unexpected seizure sometimes leads to a diagnosis, as was the case with Massachusetts Senator Edward Kennedy, who was diagnosed with glioblastoma in 2008.
Fewer than one in 50 patients survives for five years or more with a diagnosis of glioblastoma. Surgery and radiation are common treatments. Recent advances in chemotherapy have led to a few months’ improvement in survival, but chemotherapy is not generally regarded as curative. Many experimental treatments are under development, based on a growing understanding of abnormalities identified in brain tumors.
None of the DNA variants identified by the UCSF and Mayo Clinic researchers is in a part of a gene that encodes protein. The variants, however, are likely to be involved in regulation of important genes, the researchers say. The next stage of research is functional genomics – exploring biological mechanisms that might explain the association with elevated cancer risk. “Now we have something that’s real,” Wiencke says. “We know these variants affect risk, but how?” The answers may eventually lead to new strategies for treatment and possibly prevention.
John Wiencke, PhD
Wiencke, who leads laboratory research for the scientific team, notes new areas of research the group hopes to explore further. The chromosome 9 risk variant identified by the UCSF researchers is in a neighborhood crowded with risk variants for other diseases. On one side are genetic variants associated with diabetes and heart disease. On the other is a familiar cancer-associated gene called p16. But Wiencke is confident that the variants their team has discovered are more closely associated with another gene, called p15, and are likely to play a role in controlling that gene’s activity. In the cancer research community, Wiencke says, “p15 is a well-known gene, but it’s not as firmly implanted in people’s minds as being as important in cancer as p16.”
The UCSF researchers previously found preliminary evidence that glioma risk is related to immune response, including allergic responses. Wiencke notes that the chromosome 9 DNA also appears to correspond to a binding site for a molecule that affects immune responses. The high-risk and low-risk genetic variants on chromosome 20 are within noncoding regions of a gene called RTEL1. In mice, the gene is known to control the length of DNA on telomeres, the ends of chromosomes. Loss of RTEL1 is associated with chromosomal abnormalities similar to those that occur in cancer. In fact, Wiencke suggests that all of the risk variants for brain tumors identified by the UCSF and Mayo Clinic researchers in the current study appear to have some connection to telomerase, an enzyme that adds DNA to telomeres and that seems to play a role in making cancer cells immortal.
To conduct the study, the UCSF researchers scanned nearly 276,000 markers to track variations among individuals across the entire human genome. Included were 692 patients from the San Francisco study led by Wrensch, as well as another 70 patients whose DNA and clinical data are included in an NCI database. A control group of cancer-free patients with otherwise similar characteristics was included for comparison. The Mayo Clinic study participants included 174 patients and 176 controls. The UCSF researchers initially identified 13 DNA markers that appeared to be associated with glioma risk. The Mayo Clinic researchers replicated the findings for three of the 13 genetic variants in the smaller group. Leading scientific journals now routinely require these kinds of data to demonstrate replication of the initial findings from genome-wide association studies. Other collaborators from UCSF included biostatisticians Ru-Fang Yeh, PhD, and Yuanyuan Xiao, PhD; bioinformaticians Ivan Smirnov and Alexander Pico, PhD; clinical collaborators Mitchel Berger, MD, Susan Chang, MD, Michael Prados, MD, and Tarik Tihan, MD, PhD; and epidemiologists Joe Wiemels, PhD, Terri Rice, Lucie McCoy, and Joe Patoka.
Variants in the CDKN2B and RTEL1 Regions Are Associated with High-Grade Glioma Susceptibility
Margaret Wrensch, Robert B. Jenkins, Jeffrey S. Chang, Ru-Fang Yeh, Yuanyuan Xiao, Paul A. Decker, Karla V. Ballman, Mitchel Berger, Jan C. Buckner, Susan Chang, Caterina Giannini, Chandralekha Halder, Thomas M. Kollmeyer, Matthew L. Kosel, Daniel H. LaChance, Lucie McCoy, Brian O’Neill, Joe Patoka, Alexander R. Pico, Michael Prados, Charles Quesenberry, Terri Rice, Amanda Rynearson, Ivan Smirnov, Tarik Tihan, Joe Wiemels, Ping Yang and John K. Wiencke
(published online July 5, 2009)
- Epidemiology and Molecular Pathology of Glioma
Judith A. Schwartzbaum, James. L. Fisher, Kenneth D. Aldape and Margaret Wrensch
Nature Clinical Practice Neurology 2(9):494-503 (September 2006)