Gene Profiling May Predict Breast Cancer Spread

Feb. 17, 2005 -- A 76-gene "signature" has been linked with breast cancer breast cancer tumors that are more likely to spread to other organs. The new finding could mean fewer women will get chemotherapy when they don't need it, researchers say. However, others say the puzzle is not yet solved.

Currently, there are no reliable tools to predict which breast cancer patients are most likely to have cancer recurrences, writes researcher Yixin Wang, PhD, with Veridex LLC in San Diego. Wang's paper appears in the latest issue of The Lancet.

Upwards of 70% of breast cancer patients whose cancer hasn't spread to lymph nodes are successfully treated with surgery and radiation therapy, Wang writes. Yet treatment guidelines for these patients recommend that 85% to 90% also get chemotherapy to help eliminate cancer cells that might be in the body. Because there is no reliable way to determine which early breast cancer (cancer that has not spread to other organs) will spread, many women at this stage of cancer will receive chemotherapy even though it may be unnecessary.

If they can more accurately identify patients with less risk of recurrence, doctors can avoid prescribing unnecessary treatment or choose less-aggressive therapies for these patients.

Wang's study involved 286 patients whose breast cancers were localized only to the breast (lymph node-negative). None of the patients got chemotherapy after surgery. All the women's tumors were given genetic testing.

The women were followed for an average of eight years. During that time one-third of the women developed cancer recurrence.

The study showed that a set of 76 genes (gene signature) could accurately predict women at high risk of recurrence.

The gene signature was highly informative in predicting which women would have a recurrence of cancer within five years. Women with the gene signature had an almost fivefold risk of developing cancer recurrence even after taking into account other factors that traditionally predict risk of recurrence, including tumor size, a woman's age, and estrogen receptor status of the tumor.

The gene signature "could provide a powerful tool to identify those patients at low risk, preventing overtreatment in substantial numbers of patients," writes Wang.

Wang's finding "is not by itself sufficient" in identifying which women might get breast cancer recurrences, writes Tor-Kristian Jenssen, a tumor biologist with PubGene AS in Vinderen, Norway, in an accompanying commentary.

Several large studies have identified signature gene patterns to predict whether a patient's breast cancer will spread or not. However, each group of researchers came up with vastly different gene patterns.

"We are left with obvious questions of which to trust and why they differ," writes Jenssen. "The signature is there, but it is still necessary to read the fine print."