"If we can put a man on the moon, why can't we find a cure for cancer?"
It's a familiar question that embodies a collective frustration Americans have over the lack of a cure for cancer. Few diseases touch as many lives, and no diagnosis is more dreaded.
This year alone, the American Cancer Society estimates that more than 1.6 million people in the U.S. will be diagnosed with cancer and roughly half a million will die from the disease.
Despite how these statistics are reported, scientists now realize that cancer is not one disease. There are types, like breast or prostate cancer, and there are subtypes of those. Tumor cells from people with the same subtype might have different genetic fingerprints. Likewise, a person's tumor itself is made up of different cell types.
With so much variation, can we ever hope to win the war President Nixon declared on cancer more than 40 years ago?
If we continue to treat cancer simply based on its organ or tissue of origin, the answer is "probably not." (Scientists are now learning that there is a subtype of breast cancer whose tumor cells more closely resemble those of a subtype of lung cancer.) If we can tailor treatments based on the genetic fingerprint of an individual's particular type of cancer, then just maybe the answer will be "yes."
Thanks to the completion of the Human Genome Project in 2003 (with help from UC Santa Cruz researchers) and faster, cheaper DNA sequencing technology, the goal of personalizing cancer diagnosis and treatment finally seems within reach.
A catalog of defects
To make this dream a reality, scientists are creating The Cancer Genome Atlas (TCGA), a catalog of the genetic defects that lead to each particular type of cancer.
"Cancer is a disease that is caused by mutations in cells that make them grow in an uncontrolled fashion," said David Haussler, a UC Santa Cruz distinguished professor of biomolecular engineering and a member of TCGA Research Network's central coordinating committee. "Because cancer is a genetic disease, it is fundamentally important to understand what those mutations are. We have not had the technology to do that on a large scale ever before."
TCGA is a comprehensive, collaborative effort funded by the National Cancer Institute and the National Human Genome Research Institute, both part of the National Institutes of Health. It involves 150 researchers at more than two dozen institutions nationwide, including UCSC. They have begun with the 20 most common cancers and recently sequenced their 1,000th tumor genome.
Haussler and his colleagues are hoping that, like women who test positive for the HER2 gene amplification and are treated with the drug Herceptin, all cancer patients will one day receive treatments that target their particular cancer. They predict hat TCGA will transform the practice of cancer medicine and our understanding of the basics of cancer biology.
"TCGA is the largest cancer genomics project in the world. There is nothing like it," said Christopher Benz, professor and program director, Buck Institute for Research on Aging in Novato.
Benz spent most of his career at UC San Francisco, where he set up its first laboratory dedicated to the study of human breast cancers, and where he continues to treat breast cancer patients. Now, along with Haussler, he is co-principal investigator of TCGA's UCSC-Buck Institute Genome Data Analysis Center, one of seven such centers in the network.
"Together, we have a unique synergy of expertise," he said of his collaboration with UCSC partners. His job is to help identify genomic changes that are clinically relevant. "When we put our heads together, we can come to the right conclusions about what is making these cancers tick and how we might go about attacking them with our therapeutic arsenal."
While Benz brings a clinical perspective to the team, Haussler's expertise is in the application of computer science and information technology to the field of biology and medicine--a field called bioinformatics. Haussler's team built the computer methods to assemble the first working draft of the human genome. They also created the UCSC Genome Browser, an open-source, online tool used by geneticists, molecular biologists, and physicians as well as students and teachers of evolution for access to genomic information.
Today, Haussler and UCSC colleagues are pioneering new methods in bioinformatics for TCGA so that researchers can transfer, access, and store the 300 gigabytes of data that are generated for each tumor genome.
"Just moving these files from one institution to another overwhelms the standard Internet," Haussler explained. "We are building the infrastructure that will allow personalized medicine in general, and cancer treatment in particular, to become a reality," said Haussler, who is also a Howard Hughes Medical Institute investigator.
UCSC's TCGA team is also coming up with new ways of analyzing the data they and others are generating. "One of the big difficulties of translating cancer genomics to the clinic is that we can read mutations, but we can't understand them," Haussler said.
To that end, UCSC's Josh Stuart spent a year writing code for software that searches tumor genomic data for only those combinations of mutations that are biologically relevant--that is, ones that might make a normal cell into a cancer cell.
"For years, biologists have been amassing knowledge of how genes act together in so-called genetic pathways. Recently, people have begun to make databases of this knowledge," said Stuart, associate professor of biomolecular engineering. "Our contribution is that we came up with a computer program that uses data from a patient sample to figure out which genetic pathways are altered from their normal function."
A test case
Once they have identified the mutations that may be leading to a particular cancer, researchers then hope to identify existing drugs or develop new ones that interfere with the cellular or molecular changes brought about by the mutations.
Stuart, Haussler, Benz, and their TCGA collaborators put this new program to the test when they looked at data from 316 patients with ovarian cancer. Published in Nature in June, the TCGA study revealed that a particular pathway was characteristically disrupted in the tumors of ovarian cancer patients.
"Until we looked at the data in this way, no one had really appreciated the importance of that particular pathway," Stuart explained. Further analysis supported the existence of four distinct subtypes of the disease and revealed 68 genes that could be targeted by existing Food and Drug Administration–approved or experimental therapeutic compounds—most already approved for the treatment of other diseases.
Back to basics
This kind of genomic approach to cancer has not only resulted in surprising findings about individual cancers, but also promises to influence the direction of basic cancer research.
UCSC's Jeremy Sanford is a molecular biologist who studies how gene regulators, called RNA binding proteins, work. Sanford was awarded a grant by the Santa Cruz Cancer Benefit Group to study the activity of an RNA binding protein that is associated with advanced pancreatic cancer.
"We want to identify the RNA targets of the protein, understand what pathways are influenced by it, and look at how removing the protein affects the biology of the cancer cell," said Sanford, an assistant professor of molecular, cell, and developmental (MCD) biology.
Sanford said his research is guided by the data being generated by Haussler and his colleagues. "They identify variation. That leads to testable hypotheses." It's up to scientists like himself, he said, to test those hypotheses.
And just as basic scientists look to TCGA for clues to how normal cells work, so does the future of personalized medicine depend on the progress of basic science, said Doug Kellogg, a professor of MCD biology.
Kellogg studies the molecular mechanisms that control cell division, mechanisms scientists do not fully understand yet. "We really need to identify the mechanisms that control cell division," Kellogg said. "We need ways to attack tumors, and that requires more basic biology."
Revolutionizing cancer medicine
In addition to more basic biology at the bench, the genomic era is also going to require changes at the bedside, Benz said.
These changes include--though may not be limited to--a new way of assigning patients to arms of clinical trials based on their cancer genomics and new ways to find patients for those trials that does not compromise privacy.
"We need the equivalent of a transplant registry for tumor samples. We have to get organized, and we have to break down barriers so that we can disseminate information freely around the country," said Benz.
Testing treatments for different subtypes of cancer will also be a challenge, Haussler said. "The existing system for clinical trials only tests one thing at a time. The process is too cumbersome. TCGA proves that nearly every tumor has a different combination of mutations. There are far too many combinations to test them one at a time."
TCGA represents nothing less than a revolution in medicine. "Big revolutions like this create stresses and strains," Haussler said.
But the changes that are needed to make personalized cancer treatment the rule and not the exception will be worth it, Haussler predicted.
"We can imagine a world where we can come up with a treatment for any combination of mutations in a person's tumor," he said. "We have to make that world a reality."