Test may help predict colon cancer recurrence risk

May 14, 2009 By MARILYNN MARCHIONE , AP Medical Writer

A new gene test may help predict which colon cancer patients are at higher or lower risk of having their cancer return after surgery, doctors report, but whether it is useful enough to justify its likely high price remains to be seen.

The experimental test mirrors the Oncotype DX test widely used now to predict recurrence risk in breast cancer patients and whether they are more likely to benefit from chemotherapy or hormone treatments.

The new test is aimed at the nearly 40,000 Americans each year diagnosed with colon tumors that are large but have not spread beyond the bowel. The vast majority are cured by surgery alone, but there's no sure way to tell who won't be.

Chemotherapy improves survival only modestly for this group as a whole, so doctors hesitate to recommend it - only about 30 percent of patients get it now.

"We're probably not giving the therapy to all of the appropriate people and we're giving it to some people who will experience little benefit," said Dr. Durado Brooks of the American Cancer Society.

A test that could sort out who needs it most "would be extraordinarily helpful," he said.

A California-based company, Genomic Health Inc., developed an 18-gene test using tumor samples from thousands of patients in previous studies, then looked to see if it could predict risk in nearly 1,500 other patients.

The test gave individual odds ranging from about 10 percent to about 25 percent of having a recurrence in the next three years, said study leader Dr. David Kerr of the University of Oxford in Great Britain. He consults for the company, and many of the other study leaders work for or own stock in it.

That separation of risk is "not quite as good as we might like," but still gives patients one more bit of information to ponder in deciding whether to have chemo, said Dr. Richard Schilsky, a cancer specialist at the University of Chicago and president of the American Society of Clinical Oncology.

However, a second part of the study found the test couldn't predict which patients would benefit from chemotherapy. That means the test only predicts recurrence risk without informing what to do about it.

"In order for it to be clinically useful, you would want those to be in sync," Brooks said.

The company will drop the six treatment prediction genes from the test it hopes to start selling early next year, said Genomic Health's chief medical officer, Steve Shak. A price has not been set for the new test, but the breast cancer one costs $3,820.

That's pricey, but many insurers pay it because it can help avoid even costlier chemo, which can run $40,000 to $50,000 for six months for a colon cancer patient, doctors said.

Study results were released Thursday by the American Society of Clinical Oncology and will be presented at the group's annual meeting later this month.

Kerr said researchers plan more work to try to refine the test into a better predictor.

"It is a first step," said Dr. Howard Hochster, colon cancer chief at New York University and a member of the oncology group's program committee that chose the study for presentation at the meeting.

"Perhaps it's going to be most helpful in saying who does not need chemotherapy," he said.

---

On the Net:

Oncology group: http://www.asco.org

©2009 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.