Using microarrays to compare gene expression profiles in different breast cancer tumors, researchers from the Netherlands Cancer Institute and Rosetta Inpharmatics recently nailed down 70 marker genes that can help predict the clinical outcome of breast cancer.
In the study, published in this week’s issue of Nature, researchers initially studied gene expression in 98 primary tumors. The tumors included a mixture of sporadic ones and those caused by germline mutations from Dutch women under the age of 55 diagnosed with lymph node negative breast cancer. Using Rosetta’s microarrays, which each included probes for approximately 25,000 human genes, they found that about 5,000 were significantly regulated in these samples.
The researchers applied an unsupervised hierarchical clustering algorithm to the data, which divided the tumors into two large groups, one containing mostly women who later developed metastases, the other one including women who did not. Analyzing only the 78 sporadic tumors of the set with a three-step supervised classification method, the scientists found the set of 70 marker genes that accurately predicted later metastasis. They tested these marker genes on another set of 19 sporadic tumors and confirmed their predictive power, which was superior to traditional prognostic factors like tumor size, tumor grade, patient age or estrogen receptor status.
In the future, these markers might help doctors decide who should receive chemotherapy or hormonal therapy after surgery. “Based on the expression profile of the primary tumor we could evaluate the long-term follow-up of these patients,” said Laura van’t Veer, the lead author of the study.
At the moment, up to 90 percent of young lymph node negative breast cancer patients receive the extra treatment, even though only a third or fewer require it. “But since you simply don’t know who will develop a metastasis, to be on the safe side, all of them are treated,” said van’t Veer. A more accurate prognosis tool could prevent such unnecessary treatment, including the side effects and costs.
Van’t Veer’s group is validating the 70 marker genes in a group of 300 unselected breast cancer patients with both positive and negative lymph nodes but no metastases at the time of diagnosis, and wants to publish the results by the end of the year. Before the assay can enter the clinic, van’t Veer said, an independent group needs to confirm their results, and the assay has to be transformed into a routine diagnostic test. “It took us more than half a year to produce all these data,” she said, “and that is of course too long if a woman is diagnosed with breast cancer and you want to make a treatment decision based on an expression profile.”
Rosetta and the Netherlands Cancer Institute jointly filed a patent application for the profile of the 70 marker genes and will likely commercialize the assay in collaboration with another company. “We expect that within two years, we will be able to offer this test to hospitals,” van’t Veer said. --JK