Cancer has long been defined by where it is found in the body. Patients have breast cancer, prostate cancer, lung cancer.
In a significant shift, researchers are coming to believe that cancer comprises hundreds of subgroups based more on genetic makeup than location.
This new thinking means that instead of focusing cancer treatment on organs, the emphasis increasingly is going to be on finding the specific genetic changes driving an individual's cancer and targeting them with drugs. Some people who have cancers in different organs may end up taking the same drugs because common genetic mutations are involved.
Already, researchers have identified a number of genetic variations, resulting in some of the most exciting breakthroughs in oncology: Doctors now believe that the drugs Tarceva and Iressa work best in a certain subgroup of patients with lung cancer, Gleevec in particular subtypes of leukemia and sarcoma patients, and Herceptin in certain breast-cancer patients. Cancer researchers also continue to announce discoveries of new cancer subtypes in liver, brain and prostate cancer, among others.
Such subtyping is a very different approach from the way doctors currently classify cancer. In order to make a diagnosis, pathologists examine the shape and pattern of cells under a microscope, searching for patterns that are associated with broad subcategories (such as non-small-cell lung cancer). But now, researchers are using more sophisticated technologies to explore each cancer's unique gene pattern. With advances such as sequencing the DNA of a tumor, they can look for changes or mutations that might then be targeted with drugs. The goal is to develop tests to identify such variations in patients, and treat them, no matter what organ the cancer is in.
The National Cancer Institute and the National Human Genome Research Institute, both part of the National Institutes of Health, are spending $100 million over three years to do a pilot project of the Cancer Genome Atlas, which will attempt to characterize genetic mutations in cancer. The Institute of Medicine just set up a committee to explore clinical applications of these and other discoveries. The group is expected to produce a consensus report with recommendations next year.
More trials are opening based on a cancer's genetic subtype -- such as studies of how a gene known as MGMT may affect response to a treatment for a brain cancer. Also, commercial tests are available to identify mutations, such as a $975 test from Genzyme Genetics, a division of Genzyme Corp., to see if patients have a genetic mutation that makes them likely to respond to Tarceva or Iressa. Different tests are available from the Molecular Profiling Institute in Phoenix and others. And more are being developed.
The new approach involves a number of challenges. In the case of some of the new discoveries, there aren't drugs yet to target the subtypes, meaning the new information won't yet make a difference in treatment. The technologies that make subtyping possible, such as techniques for examining a thousand genes in a cancer all at once, are costly, and it will require large sums to get information on every cancer.
There also are concerns that researchers would spend large amounts of time and money developing drugs that might help only a small group of people. Even in more common cancers such as lung cancer or breast cancer, just a fraction of patients may share common genetic mutations and respond to drugs targeted against them.
Some researchers, including Bruce E. Johnson, director of the Carole M. and Philip L. Lowe Center for Thoracic Oncology at the Dana-Farber Cancer Institute in Boston, argue that by targeting subgroups rather than searching for a cure for everyone that has so far proven elusive, doctors may have a better chance of ultimately curing more people. The idea isn't to treat everyone the same and end up with drugs that make many people "only 20 percent better," says Dr. Johnson, one of the patent-holders on a test to subdivide lung-cancer patients. "Instead, we'd like to start making 20 percent of the people 100 percent better."
It is also possible that because a genetic variation may cut across various organs, a tailored drug could be used more broadly. For instance, Novartis AG's Gleevec is approved to treat subgroups of two different cancers, gastrointestinal stromal tumors and chronic myelogenous leukemia. Now Novartis says Gleevec is also effective for four other diseases, including certain blood cancers and a kind of skin tumor.
For some patients, subtyping already is making a difference. In July 2004, Marguerite Blattner, now 53, was diagnosed with stage 4 lung cancer. As a female who had never smoked, the La Jolla, Calif., woman had characteristics that could make a difference in her prognosis. Still, she found she was "thrown into this big pool where you're just one of hundreds of thousands of people with lung cancer. You're not differentiated in any way by who you are," she says.
She saw Dr. Johnson at Dana-Farber, who recommended that she get her tumor tested to see if she had a mutation that would make her likely to respond to Tarceva or Iressa, which have far fewer side effects than chemotherapy because they target only cancer cells, not normal ones.
"The doctor called me and said, 'Congratulations, you have the mutation,' " Ms. Blattner says. Instead of doing standard chemotherapy first, Ms. Blattner quickly went on Iressa. She still is on the drug and has had such minimal side effects, she said, "that I can't even categorize myself like other lung-cancer patients who get chemo."
Last year, in a paper published in the New England Journal of Medicine, researchers announced they were starting to understand why some patients with glioblastoma multiforme, a brain cancer, are likely to live longer if they are treated with the drug temozolomide. One of the keys was a gene known as MGMT, which is involved with DNA repair. In 60 patients in whom the gene was active, two-year survival was just 13.8 percent after treatment with temozolomide and radiation. For 46 patients in whom the gene was "silent," treatment was much more effective, and two-year survival was 46 percent. The trial was organized in seven countries by the European Organization for Research and Treatment of Cancer and the National Cancer Institute of Canada Clinical Trials Group.
Schering-Plough Corp., the maker of temozolomide, and Oncomethylome Sciences are now developing technology to measure the status of the MGMT gene in patients with glioblastoma. There are new trials targeting patients with the silenced gene, as well as for patients with an active MGMT gene to find ways to overcome their resistance to the drug. Some groups are also considering research into whether the MGMT gene plays a role in lung-cancer patients.
Some researchers have questioned the move toward launching trials open only to people with particular subtypes -- arguing that trials open to people with cancer in the same organ was how smaller groups were first discovered. The subgroups were ones who didn't respond like the other patients.
Warren Mason, a neuro-oncologist at Princess Margaret Hospital in Toronto and one of the co-authors of the glioblastoma paper, says that such trials don't necessarily mean that other patients with the same disease shouldn't have access to a targeted drug. "I don't think there is enough evidence to say that some people with glioblastoma should not get temozolomide," he says. "It's the only drug we have that's effective for it, and even if it is only helping some people a little, that's important in a lethal cancer."
One of the most frustrating aspects of research into subtypes is the lack of drugs to target what researchers are finding, says Bert Vogelstein, an investigator at the Howard Hughes Medical Institute and director of the Ludwig Center for Cancer Genetics and Therapeutics at Johns Hopkins University.
Dr. Vogelstein's lab found that a particular gene called PIK3CA was mutated in about one-third of colon cancers. The mutation increases the activity of an enzyme, and drugs could theoretically be developed to inhibit the enzymes. Investigators at other institutions have found that the mutation also appears in a "significant fraction" of brain, ovarian, breast and liver cancers, says Dr. Vogelstein, meaning that "it's a perfect target for drug development."
Still, turning a tantalizing discovery into a useful drug "could take 10 years, and that's being optimistic," he says.