Subscribe for 17¢ / day

CHICAGO — Laboratory tests routinely used to determine if breast cancer patients should receive a life-saving new drug often yield false results, so some women who desperately need the drug do not receive it, and others who take it unnecessarily are exposed to potentially dangerous side effects.

According to several recent studies, up to 18 percent of the tests resulted in false-positive errors, meaning women who have little chance of benefiting from the costly drug were mistakenly classified as eligible.

Because the tests also suffer from false negatives, some patients for whom the drug might be the last, best chance are denied access. Experts disagree on the extent of the false negatives, but according to one researcher, as many as 20 percent of the women who might be eligible to receive the drug test negative.

The problem could affect tens of thousands of people. More than 2 million U.S. women have breast cancer, and an estimated 192,000 new cases will be diagnosed this year.

The drug, Herceptin, manufactured by Genentech, is an antibody approved by the Food and Drug Administration less than three years ago for advanced breast cancer patients whose disease had metastasized, or spread to other organs. An estimated 15 to 20 percent of breast cancer cases metastasize, and such patients are generally considered incurable, but Herceptin prolonged the lives of many of them in clinical trials.

Giving the drug to the wrong patients is problematic because it appears that only those whose tumors contain large numbers of a cell-surface protein called HER2 stand to benefit from it. But everyone can experience the drug’s possible side effects, which include a higher risk of heart failure, especially when the drug is given along with certain chemotherapy agents.

Dr. Ken Bloom, a pathologist at Rush-Presbyterian-St. Luke’s Medical Center in Chicago, said a standard course of Herceptin therapy costs $50,000 to $60,000 a year, and some patients continue to get the weekly infusion for years.

“No one who was truly HER2-negative responded to the drug (in clinical trials), so it’s a big waste,” Bloom said.

Herceptin neutralizes HER2, a receptor whose function is to transmit growth signals to the cell’s nucleus. Normal breast cells contain a small amount of HER2, but when the protein is overabundant, the cells can grow and multiply out of control, forming cancerous tumors.

Newly diagnosed breast cancer patients are routinely tested for HER2. The results help doctors predict how aggressive the cancer is, what type of chemotherapy might be best and whether the patient is likely to respond to Herceptin.

Patients with an excess of HER2 have a particularly bleak prognosis. Their cancers often grow fast, have a tendency to recur and metastasize, and may be resistant to some standard cancer drugs.

Herceptin, however, is much more likely to work in patients whose tumor cells are strongly positive for HER2 than in those who have only moderately increased proteins.

Experts estimate that about 20 percent of breast cancer patients are HER2-positive; of those, 20 to 40 percent can be expected to respond to Herceptin.

One of the lucky ones is Jean Rachell, 69, whose breast cancer — diagnosed in 1986 — spread to her liver, lungs and bones. Before Herceptin, Rachell’s only option would have been toxic chemicals. But she has been taking the new drug for about a year now, with good results.

“I feel great,” said Rachell, who is being treated at the University of Chicago. “It’s nothing like being on chemotherapy — and I feel like it’s keeping the cancer at bay.”

Problems with testing for HER2 have emerged as researchers study whether the drug is also useful for women in earlier stages of breast cancer. As they double-check their subjects to ensure accurate results, they have found that an alarming number of them had been mistakenly identified as eligible for Herceptin.

The scientists say most of the bad results occurred when less-experienced laboratories processed the samples. Others debate whether one type of test might be more accurate than another.

When the FDA approved Herceptin in 1998, the agency also approved a commercial test kit called HercepTest, made by Dako, which uses an antibody-based stain to determine whether a tissue sample is positive for HER2. Three other tests for HER2, including one other antibody assay, also have been approved.

In an antibody assay, each sample is scored as 0, 1+, 2+ or 3+, and tumors deemed 2+ or 3+ are considered HER2-positive.

But researchers noticed early on that the response rates were much higher in the 3+ patients than in those with a score of 2+. As a result, several ongoing clinical trials in which Herceptin is being tested require a 3+ score for eligibility.

Even that precaution, however, may not be enough to screen out women who are not likely to benefit from the drug.

The National Surgical Adjuvant Breast and Bowel Project (NSABP), a research cooperative sponsoring a trial of the drug in 2,700 women with early breast cancer, retested the first 104 patients enrolled in the study as a quality-control measure and found that 19 of them were not eligible.

Nearly all the false-positives came from labs that don’t do a lot of HER2 testing or that use assays other than HercepTest.

When small hospital or community-based labs used HercepTest, 10 out of 52 patients (19 percent) turned out to be ineligible. The results were even worse when such labs performed antibody assays other than HercepTest: 8 of 23 (35 percent) were false positive.

On the other hand, “reference” labs — facilities that have processed 100 HER2 assays a month for six months — almost never got them wrong.

As a result, NSABP now requires that patients enrolled in its Herceptin trial have their eligibility confirmed by HercepTest in a reference lab or by a DNA test known as FISH, or fluorescence in situ hybridization.

The reference labs recognized by NSABP include Impath, Quest and Rush-Presbyterian-St. Luke’s.

Dako said the University of Chicago Hospitals, Alexian Brothers Medical Center in Elk Grove Village, Northwest Community Hospital in Arlington Heights, Hinsdale Hospital and Evanston Northwestern Healthcare also do high-quality work with HercepTest.

Another large trial is being run by the North Central Cancer Treatment Group. The principal investigator of that trial, Dr. Edith Perez, said the group reviewed the tests of 114 patients. Although the results have not yet been released, Perez said, “Our data are consistent with the NSABP’s.”

“We have a tremendous amount of concern,” said Perez, who also directs the breast cancer program at the Mayo Clinic in Jacksonville, Fla. “There’s an unacceptable rate of false positives based on community testing; 25 to 30 percent (of those who test positive in small labs) could be ineligible for Herceptin therapy.”

As for the advanced breast cancer patients for whom the drug is prescribed outside a clinical trial, Perez said, “Many women are getting Herceptin based on those same tests.”

While HercepTest and other antibody assays count the HER2 proteins on the cell surface, the FISH test counts the HER2 genes in the cell’s nucleus. Several recent studies have suggested that extra copies of the gene are a better measure of HER2 positivity than extra copies of the receptor protein.

But the FISH test is considerably more expensive and more time-consuming, and most labs are not equipped to do it.

Dako says many labs could improve their HercepTest results simply by following the manufacturer’s instructions.

“The NSABP found that HercepTest is the most consistent” of the antibody assays, said Martha Townsend, a Dako official. “It could be reproducible in small labs, too, if the pathologists did it correctly.”

Several physicians confirmed that many labs do not follow the Dako protocol but improvise or use their own procedures.

“You can’t do that,” Bloom said. “If you deviate, it disrupts the interpretation.”

Dako recommends screening all patients with HercepTest and then using FISH as a backup for people who score 2+ (about 10 to 15 percent of the total). That’s standard practice among many doctors because most false positives are 2+.

But Dr. Robert Mass of Genentech said any antibody test that is less than 3+ should be redone with FISH to avoid the problem of false negatives. According to his research, 20 percent of patients who were found to have extra copies of the HER2 gene had scored negative (0 or 1+) on the antibody test used in clinical trials.

“If the goal is to identify all those patients who are likely to benefit (from Herceptin), then false negatives are a much bigger problem” than false positives, he said.

A Genentech official said about 35,000 women have taken Herceptin since it was approved.

Copyright © 2001 Chicago Tribune. Distributed by Knight Ridder /Tribune Information Services. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.

Subscribe to Breaking News

* I understand and agree that registration on or use of this site constitutes agreement to its user agreement and privacy policy.