By Malorye A. Branca
March 10, 2003 | Agendia, a spin-off from the Netherlands Cancer Institute, plans to commercialize the first DNA chip-based test, which aims to predict which women have the most aggressive breast tumors. It was developed by Institute researchers in collaboration with staff at Rosetta Inpharmatics Inc., acquired by Merck and Co. in 2001, and will debut in Europe.
“In Europe, doctors and patients are more willing to try new things,” says Laura J. van’t Veer of the Netherlands Cancer Institute, one of the test’s developers. “The regulatory environment is more open than in the United States.”
Probably the major challenge will be convincing reimbursement agencies that the test is worth paying for. The research group is in discussions with Dutch authorities about coverage. The first target market is the Netherlands, where one hospital will soon start using the test experimentally. “It is done in a quality-certified laboratory,” van’t Veer says. “That certification is recognized throughout Europe, but there are different reimbursement agencies.”
If Agendia’s plan works, the test could be commercialized by year’s end. The company plans to follow up with other gene-expression-based tests.
Last year, the Dutch group and its collaborators grabbed headlines when it showed that a gene-expression signature, or pattern, was highly accurate at distinguishing breast tumors most likely to spread from less dangerous ones. Such a test could help doctors decide which women most need follow-up drug therapy after breast cancer surgery.
That study looked at about 80 banked tumor samples from women whose final diagnosis was already known. In a more recent study, the group ran the same test against almost four times as many tumors. Those results appeared in the December issue of the New England Journal of Medicine. The study represent a milestone because many supposed gene-expression markers have been found, but few have been shown to work on more than one set of samples.
The Dutch researchers are now launching two large-scale trials with about 5,000 patients each. In these studies, the test will be done on samples before the diagnosis is known. One study will run in the United States, and the other will take place in Europe with the European Organization for Research and Treatment of Cancer. Final results are expected in about four years.
A New Window into Breast Cancer
Cancers can act differently even when they look the same. To decide how to treat breast tumors, doctors look at a wide range of indicators such as whether the cancer has spread to nearby lymph nodes, how big the tumor is, and certain characteristics of the tumor cells. However, none of these factors is very accurate.
The DNA-chip test, meanwhile, reveals how 70 genes are turned on or off in the cancer cells. According to the Dutch study, the tumors most likely to spread usually show a different pattern of gene expression than their less dangerous counterparts.
Most importantly, the test appears to be more useful than any information doctors already use. “Nothing else was better at predicting a woman’s risk of metastases [the spread of cancer to different parts of the body] than this gene signature,” van’t Veer says.
Some experts, particularly in the United States, have questioned whether a DNA chip can be a practical clinical test.
“I think it remains a possibility that some sort of microarrays will make it into the clinic,” says Todd Golub, director of cancer genomics at the Whitehead Institute Center for Genome Research and an oncologist at the Dana-Farber Cancer Institute in Boston. “I’m not sure how they would get approved, given their complexity, and I’d want to see them very well validated.”
“It’s extremely difficult to reproduce DNA-chip results even in your own lab,” said Nicholas Dracopoli of Bristol-Myers Squibb, speaking at the American Association for Cancer Research’s recent “Oncogenomics 2003” conference. “It can be done but it takes a lot of experience.”
The Dutch group is confident in its test’s efficacy. “The first two years we did have difficulties with reproducibility, but now we are seeing very consistent results with the chips,” van’t Veer says. To simplify it, the version of the test that hits the clinic may contain fewer than 70 gene markers.