December 11, 2004
Data Mounts on Avoiding Chemotherapy
By ANDREW POLLACK
New findings add to the evidence that genetic tests can help predict whether breast cancer will recur, giving valuable guidance to doctors and patients about whether potentially toxic chemotherapy will be useful or can safely be avoided, researchers said yesterday.
The findings, presented at a breast cancer conference in San Antonio and in a paper released by The New England Journal of Medicine, suggest that telltale genetic signatures of tumors will in time be used to tailor treatments to individual patients.
"It would probably make a difference in the treatment of tens of thousand of women," said Dr. Robert C. Bast Jr., vice president at the M. D. Anderson Cancer Center in Houston and co-author of an accompanying editorial released by the journal.
But Dr. Bast added that the tests were still "a work in progress" and needed more study before they could be used routinely.
The tests address a quandary for patients with breast cancer. When tumors are detected early and removed surgically, many women undergo chemotherapy to lower the chance that the cancer will recur.
But for a vast majority of women, the cancer will not recur regardless of whether they receive chemotherapy. So they are exposed needlessly to the treatment, which can cause nausea, hair loss, vulnerability to infections and, more rarely, cardiac problems or leukemia. Doctors cannot now tell, however, which women need the chemotherapy.
"There's a substantial portion who we either undertreat or overtreat because we don't have adequate information on who will recur," said Dr. Gary H. Lyman of the University of Rochester, who has reviewed 10 such experimental tests.
The new tests "enhance the ability to distinguish between low risk and high risk," said Dr. Lyman, who was hired to do a cost benefit analysis for the maker of the test described in the paper released yesterday.
That test was developed by Genomic Health, a company in Redwood City, Calif. The test looks at the activity levels of 21 genes in tumor samples. It categorizes the cancers as being of low, intermediate or high risk of recurrence.
To validate the test, company scientists and academic collaborators from the National Surgical Adjuvant Breast and Bowel Project used samples from 668 women. Those women had taken part in a trial in the 1980's, so it was already known whether their cancers had recurred.
Only 6.8 percent of the women rated by the test to be in the low-risk group suffered a recurrence outside the breast within 10 years, compared with 14.3 percent in the intermediate group and 30.5 percent in the high-risk group.
Those results apply only to women who are taking a hormonal drug called tamoxifen for newly diagnosed cancer that has not spread to the lymph nodes and that is stimulated by estrogen. Still, there are estimated to be at least 50,000 new cases a year that fit that category.
The data on risk, initially reported at the San Antonio Breast Cancer Symposium a year ago, showed only whether the cancer was likely to recur, not whether chemotherapy would help prevent a recurrence.
But new data presented in San Antonio yesterday, from a different trial by the same authors, showed that the patients in the high-risk group had benefited from chemotherapy while those in the lower-risk groups had not. In the high-risk group, 40 percent of patients who got only tamoxifen had a recurrence of cancer within 10 years, compared with just 12 percent who got both tamoxifen and chemotherapy. But in the low-risk group, both those who got chemotherapy and tamoxifen and those who got tamoxifen alone had a recurrence rate of roughly 5 percent.
Genomic Health is already selling its test, for $3,460. Since the company performs all testing in its own laboratory, it did not need approval from the Food and Drug Administration as it would if it were selling test kits to doctors.
Genomic Health is hardly the only company in the field. A Dutch company, Agendia, is offering a test that looks at 70 genes. And Exagen Diagnostics, based in Albuquerque, introduced preliminary data in San Antonio on a three-gene test it hopes to bring to market in 2006, one that it said would work for a broader range of breast tumors than the other tests. Exagen said its test's prediction that cancer would not recur was correct 91 percent of the time.
Yet a fourth company, Arcturus Bioscience, based in Mountain View, Calif., is also developing a test.
The tests vary not only by which genes they look at but by how they are performed. Agendia's requires fresh frozen tumor samples, which are not always obtained. Genomic Health's test uses samples preserved in paraffin wax, the usual procedure, which doctors said would make that test more practical. Exagen said its test would also use standard samples and could be performed by many laboratories, not just in a central location.
So far, the tests have not been widely used, as doctors await more study. Scientists said it was not clear, for instance, whether the Genomic Health test would apply to women who get newer drugs called aromatase inhibitors, which are starting to replace tamoxifen.
Insurance companies do not routinely pay for the tests.
nytimes.com |
