What You Need to Know About Mapping the Cancer Genome

Two of the most feared cancer diagnoses are glioblastoma, the most aggressive kind of brain tumor, and pancreatic cancer. Both kill the average patient within months, not years, of diagnosis, so there's special interest in finding the Achilles' heel of those cancers. Scientists are now reporting that they've gotten an up-close-and-personal look at the genetic mutations linked to those two cancers, which may—eventually—lead to better matching of appropriate treatments to individual patients, new diagnostic tests, and possibly even entirely new drugs.

Human Chromosomes

Video: Learning About Cancer

Beyond the implications for those specific cancers, there's a great deal of interest in the way the two groups of researchers approached their work: Both analyzed huge quantities of genetic information from different tumor samples in an attempt to essentially catalog the many different permutations of cancer. You're going to hear more about this approach in the future; one group, the government-funded consortium known as the Cancer Genome Atlas (TCGA) Research Network, is hoping to map even more cancers in coming years. And other researchers in the public and private sectors also are focusing on the cancer genome. Here's what you need to know:

What is the cancer genome, anyway?
Cancer, by definition, is caused by genes going haywire—mutating in ways that are not part of the normal blueprint for the body's form and function. Even among tumors of a single type, those genes can go haywire in many different ways. So describing the entire cancer genome involves characterizing all the different abnormalities that are linked to each of the 50 major types of cancer. That daunting task would involve churning through 12,500 times as much information as scientists processed in their recently completed effort to map the entire human genome.

What's the Cancer Genome Atlas project?
TCGA, as it's called, is a government-funded consortium charged with chipping away at this task. Now in a three-year pilot, it is focusing on just a few forms of cancer, including glioblastoma. (It just published a study in the journal Nature. That was one of three reports on the subject that appeared this week.) If those efforts prove promising, the full project, costing a proposed $1.5 billion and lasting many years, could get the green light. But technology has not yet made it practical to characterize every single gene in every tumor, so the TCGA's approach is to focus on the genes already known to be active in tumors. Those genes still contain about 100 times as much genetic material as was mapped in the human genome project.

Are other researchers doing the same thing?
Yes. A separate group published two papers in Science, one describing glioblastoma and the other characterizing pancreatic cancer. The two groups' aims are the same, but their approaches are slightly different. The TCGA team looked at more than 200 tumor samples but analyzed only genes already identified as active in cancer, says Steve Elledge, a geneticist at Harvard Medical School. The group publishing in Science, by contrast, looked at fewer samples but cast a broader net: It studied all the genes in those samples, not just the ones previously known to be associated with cancer. And it linked a new gene to glioblastoma.

So when will new gene-based cancer treatments be on the market?
Not soon. This is all very preliminary. The unifying conclusion of the new research is that there are many different mutations in a given type of tumor. That will make it difficult to come up with drugs targeted at every mutation. "The conclusion has subtly shifted," says Jeff Boyd, chief scientific officer of the Fox Chase Cancer Center in Philadelphia, who wasn't part of either research group. Now, he says, researchers are more likely to take aim at molecular "pathways" in a cancer cell. A pathway, in this case, is a sequence of molecular events in a cell that lead to a single end result but that may be sparked by several different types of mutations.

Think of it this way: Multiple cars might be taking their own routes to a given garage, but they're all planning to end up in the same place. Instead of trying to stop each car on the road, you could put up one roadblock in front of the garage. That promising approach to stopping cancer is attracting a lot of research attention and dollars, but there are as yet no drugs on the market targeting pathways, says Boyd. It's possible, but it would take years.

If not new treatments, then what?
More immediately, this kind of genetic mapping is likely to help characterize patients' tumors. Perhaps one type, sparked by a particular mutation, is more likely to respond to one therapy than another. Knowing that could help doctors decide which drug is the best to use for each individual patient. Also possible are new diagnostic techniques for identifying cancer earlier, though those, too, are years down the line.

What's the downside of mapping the entire cancer genome?
Some scientists question if the proposed effort is worth its anticipated price tag. At the very least, they argue that other strategies, such as looking at genes in tumors that aren't mutated but are still required for the cancer's survival, should be pursued in parallel to a cancer genome mapping effort. That's Elledge's view of things (his research lies within that alternative approach). The genome work is "bearing some fruit," he says. "The question is whether this fruit is worth the price." Christopher Logothetis is chairman of the department of genitourinary medical oncology at M. D. Anderson Cancer Center in Houston and a member of the external scientific committee for the TCGA. He says the new study was a successful pilot. "It demonstrated we could get people together, drive down the cost, find tissue, and learn something." Now, he says, it's time to figure out how to characterize the cancer genome, given limited resources, and whether it's best done by the public sector, private sector, or both.