Healing Gone Haywire: Wound-repair genes signal cancer spread

Aggressive tumors have a lot in common with wounded tissues—both show rapid cell division and the growth of new blood vessels. Those similarities inspired a new test to predict which breast tumors will spread rapidly if untreated and which are likely to be less aggressive.

The test, which tracks the activity of genes that normally mend injured tissues, outperforms existing predictors of breast cancer spread, or metastasis. Researchers say that the test could eventually enable physicians to be more selective in recommending chemotherapy to fight the spread of a cancer within the body.

The researchers assign a score to each tumor on the basis of how closely its gene activity matches that of healing tissue. “It turns out that that number is really strongly [negatively] associated with survival and time to metastasis,” says team member Trevor Hastie of Stanford University.

Previous studies looked at which genes are turned on or off in aggressive cancers versus slow-growing tumors. But most of this work, which tracks thousands of genes at a time, has failed to turn up consistent patterns of gene activity. The challenge has been to find a pattern essential to the spread of tumors.

Focusing on the similarities between aggressive tumors and wounded tissue, Howard Y. Chang of Stanford University and his team zeroed in on genes that promote healing. The researchers first identified what they call the wound-response signature, a characteristic pattern of activity among 500 genes in nonmalignant tissue recovering from wounds.

The team then used DNA microarrays, which can measure the activity of thousands of genes at a time, to discern patterns in tumor samples from an earlier study of 295 breast cancer patients at the Netherlands Cancer Institute in Amsterdam. As the team reports in a forthcoming Proceedings of the National Academy of Sciences, tumors showing the wound-response signature were seven times as likely to have spread as were those without it.

The wound-response signature test correctly identified 90 percent of the tumors that had spread in these women, and it flagged some aggressive tumors missed by a previously developed 70-gene test (SN: 02/02/02, p. 68: Available to subscribers at Ominous signals: Genes may identify the worst breast cancers). The new test also identified as low-risk 60 patients who didn’t get chemotherapy and whose cancer didn’t spread in 12 years of follow-up.

The test “can allow us to make more-informed decisions of whether chemotherapy is appropriate or not,” says Hastie. Currently, many cancer patients are given chemotherapy when they don’t actually need it, he says.

Although other scientists have found gene-activity patterns in cancers, says Mark E. Ewen of Harvard University, “none of them actually give you the underlying mechanism” that makes some tumors more aggressive. By contrast, says Ewen, Chang’s study identifies the wound response as the factor that makes some breast tumors spread.

“It tells you something about why those cancers are higher risk,” says Chang.

The group is now looking into the possibility of drugs that could shut down healing-style activity in tumors and thus prevent them from spreading. The researchers also intend to study whether patients benefit from treatment decisions based on the wound-response signature.