Colon Cancer Gene Test Can Predict Disease Return, Study Says
The test, called Oncotype DX for colon cancer, sifts through patients tumor cells for a telltale pattern of genes that indicate a two-fold increased risk of relapse, according to a study previewed yesterday by the American Society of Clinical Oncology. The test failed to predict whether chemotherapy drugs would actually work for high-risk patients.
Doctors cant absolutely predict when colon cancer, diagnosed in 146,970 Americans yearly, might return in early- stage patients. While the new test may be useful “as a tie- breaker” in patients whose need for chemotherapy is unclear, it isnt the final answer, said Alan Venook, professor of medicine at the University of California San Francisco.
“You got half a loaf, instead of a whole loaf,” said Venook, who has consulted for Genomic Health, in a telephone interview yesterday.
The study, led by David Kerr at the University of Oxford in England, is one of about 4,000 presentations planned for the American Society of Clinical Oncologys annual meeting May 29 to June 2 in Orlando, Florida.
Genomic Health of Redwood City, California, plans to launch Oncotype DX for colon cancer in early 2010, said Emily Faucette, a company spokeswoman. The company already markets a similar gene test for breast cancer, at a cost of $3,820, to gauge risk of relapse as well as predicting if they will benefit from chemotherapy.
The research summary noted lead author Kerr is a consultant and has received research funding from Genomic Health.
Shares Rise
Genomic Health rose as high as 11 percent to $21.00 in extended trading, after closing up 1 cent at $18.91 in Nasdaq Stock Market composite trading prior to release of the report.
The research confirmed that in colon cancer patients taking a new version of the Oncotype DX test, those with a particular 18-gene signature had a 25 percent risk of relapse. People who did not show the same gene pattern had a 10 percent risk that the malignancy would return, the study found.
That information is valuable to doctors who otherwise cant predict which colon cancer patients will relapse, said Richard Schilsky, president of the cancer organization, in a briefing.
“Twenty percent of cancers recur and we cant tell who the 20 percent are,” Schilsky said. “Its a big black box.” The test is “not as powerful as we would like,” he said.
Schilsky said doctors may want to use this method along with other predictive factors, including a patients age, general health, tumor stage, and willingness to get chemotherapy.
The colon cancer gene study, called Quasar, was “a landmark” because it selected patients who should consider chemotherapy after surgery, even though the study failed to meet its second goal of predicting clear benefit from such treatment, said Steven Shak, Genomic Healths chief medical officer, in a telephone interview.
Shak said the company worked almost four years to identify an 18-gene signature from a bank of tumor samples collected from more than 1,851 patients. Quasar confirmed that the 18-gene test identified recurrence risk in 1,436 patients with stage 2 colon cancer.
Failure to Predict
Use of the gene test failed to predict whether patients would get better using the standard chemotherapy drugs 5-FU and leucovorin, the study abstract concluded.
A separate study previewed yesterday by the oncology society found another genetic test can tell lung cancer patients whether to get chemotherapy or a new drug, according to Schilsky. The research by Masahiro Fukuoka in Japan found that certain gene mutations can inform patients whether their lung cancer will respond best to AstraZeneca Plcs Iressa or standard chemotherapy drugs carboplatin and paclitaxel.
“Patients showed some striking differences,” said Fukuoka, a professor of medicine at Kinki University School of Medicine in Osaka, Japan, in the telephone briefing yesterday.