College of American Pathologists
Printable Version

  Feature Story


cap today

Erbitux makes its entrance, with test in tow

April 2004
Anne Paxton

In the world of cancer treatment, the fate of patients and the fortunes of pharmaceutical companies often ride on the difference between "might help" and "probably will help." A tantalizing, though elusive, goal is finding a diagnostic test that perfectly flags patients who will benefit from a particular therapy.

Colorectal cancer patients got both a new therapy and a new diagnostic test in February when the Food and Drug Administration approved ImClone Systems’ cancer drug, Erbitux (cetuximab), a monoclonal antibody. The associated diagnostic test, EGFR PharmDx, made by DakoCytomation, was approved the same day. It will give clinicians a broad indicator of who is likely to benefit from Erbitux treatment.

"This is a breakthrough event that ushers in a new wave of targeted therapy for treatment of colon cancer," says Dennis Chenoweth, MD, PhD, DakoCytomation’s corporate vice president of diagnostics.

Clinicians are similarly upbeat about Erbitux’s prospects, says Mark Pegram, MD, associate professor of medicine at the University of California at Los Angeles Geffen School of Medicine. "For the last 30 years, there have only been a couple of drugs available to treat colon cancer, and within the past few weeks all of a sudden there are two new drugs. So it marks a new era of molecularly targeted therapeutics, and we can expect in the future for important drugs like this to be approved on a more regular basis."

"It’s extremely exciting," he says, "because now we suddenly have a reasonable number of different drugs to try, and they’re remarkably well tolerated. It’s not a chemotherapy; they’re antibodies, so they don’t cause nausea or vomiting."

Last October, hopes for advanced colon cancer patients rose with approval of Avastin, a Genentech drug that lengthened the lives of such patients by 30 percent, an average of 4.7 months. Avastin works by blocking vascular endothelial growth factor, or VEGF.

Erbitux, which targets epidermal growth factor receptor, or EGFR, is the drug at the center of the ImClone Systems insider trading scandal. In December 2001, "everyone expected Erbitux would be approved‚ but it got denied because the FDA requested more data," says Kenneth Bloom, MD, senior medical director of US Labs, Irvine, Calif.

Early rumors of the application denial led company insiders to dump their ImClone stock before its value plunged. But ImClone’s fortunes changed when it was later acquired by Bristol-Myers Squibb. Now, Erbitux has already paid off handsomely for the company, which has been paid $650 million so far by Bristol-Myers and stands to collect another $250 million soon.

DakoCytomation, a Danish company with U.S. headquarters in Carpenteria, Calif., is the world’s largest manufacturer of immunohistochemistry reagents and a "leader in the field of pharmacodiagnostics," says Dr. Chenoweth. It was at the FDA’s urging that EGFR PharmDx came to be developed, he says. "When Herceptin was screened by the FDA for treatment of breast cancer, the agency said Genentech needed a diagnostic test to see which tumors expressed a receptor that would allow the monoclonal antibody to bind, in order to target patients most likely to respond." Genentech turned to DakoCytomation, which developed the HercepTest and launched it the same day Herceptin was approved.

The company repeated the process with ImClone’s Erbitux. "ImClone thought it would be good to approach us to develop a similar diagnostic test to aid clinical trials in the selection of patients and to be sure that when the drug was approved, a diagnostic test would be available. So we’ve really acted in a co-development agreement with ImClone since 1999," Dr. Chenoweth says.

Cetuximab had been in the pipeline for some time as a therapeutic for advanced colon cancer, Dr. Pegram says, but only recently were there enough clinical data for the FDA to make a formal approval through documentation of efficacy. "Convincing data came from some European studies which confirmed North American clinical trials in showing single agent activity in advanced, heavily pretreated colon cancer and even greater activity when combined with chemotherapy like Irinotecan [CPT11], even in Irinotecan-resistant patients," he says.

The pivotal trial establishing Erbitux’s value was carried out in Germany, where patients with EGFR PharmDx-positive test results were treated with Erbitux, either alone or in combination with Irinotecan. Of 577 tumor specimens tested, approximately 82 percent were positive for EGFR PharmDx expression. The patients were randomly assigned to two types of treatment. Those treated with chemotherapy plus Erbitux achieved a response rate of 22.9 percent, and patients receiving Erbitux alone achieved a response rate of 10.8 percent.

Controversy has surrounded ImClone’s clinical trials, however. "The original study was such that they were trying to identify patients who expressed the EGFR receptor on the surface of the tumor cell, then treating them with Erbitux," Dr. Chenoweth says. "The idea was that they would have a greater opportunity to identify and treat patients likely to respond."

"But some clinicians have argued that that really is not a perfect trial design from a diagnostic standpoint," he adds. "What you’d really like is to take all patients whether they’ve expressed the receptor or not, treat them, and see what the clinical response is and correlate the clinical response to a diagnostic test. That was not done." Investigators in postmarket surveillance studies, however, will address this issue, he says.

There is already a brisk market for the test and the drug. In Switzerland, where Erbitux was approved in December 2003, "demand has been stronger than we anticipated," Dr. Chenoweth reports. US Labs guessed early this year that Erbitux would be approved and that it needed to get ready, Dr. Bloom says. "This drug is primarily used for metastatic colon cancer patients who have failed chemotherapy. That’s only a subset of colorectal cancer patients, but right now there are a large number of patients in this category who qualify for Erbitux therapy." There’s a huge demand for testing, he says, though once the first wave has been tested, the number of candidates may drop.

"We were early believers," Dr. Bloom says. "We went to DakoCytomation because of their high quality and the knowledge that they had the PharmDx kit in development. We are one of their large customers, and we told them we wanted as much experience as we could because it was going to take off."

"Anticipating approval, we preordered kits so that we were ready to roll when the FDA approval was announced. We’ve been running about 60 cases a day, which is significantly higher than the 800 tests per month that we anticipated."

EGFR is in a class of receptors found on tumor cells of breast, lung, colorectal, and kidney origin, as well as several other forms of cancer. Preclinical studies showed that EGFR inhibitors could sometimes bring tumor growth under control. "Researchers grew tumors in animal models, and when they inhibited the expression of EGFR or blocked the EGFR pathway, the tumors would shrink or stay stable," Dr. Bloom explains. "These data led several companies to target the EGFR pathway."

Says Dr. Chenoweth, "In more and more studies by academic investigators, pharmaceutical companies, and diagnostic companies like ours, lots of people are beginning to unravel the mysteries of what is called translational medicine—the idea of looking at the signals that cause cancer cells to grow and proliferate and metastasize."

"It’s becoming apparent," he continues, "that it’s not just one receptor like EGFR, but a combination of these receptors that provide the signals. So while the original idea might have been a ’magic bullet’—a single drug working a single receptor-we’re now looking at combinations of drugs working together and combinations of diagnostic tests. For example, we might combine HercepTest and the EGFR test and look for cells that express both receptors at the same time, and hypothesize that those would be the cells most responsive to therapy."

DakoCytomation has other products under development that it hopes will similarly sharpen clinicians’ ability to save cancer patients from treatments that won’t work. "If you had a diagnostic test where you could look at a tumor sample and say, ’This patient does not express the target receptor,’ you could not only save money but spare the patient exposure to a drug," Dr. Chenoweth notes.

One test for which the company has completed clinical trials and is now preparing an FDA submission relates to the breast cancer chemotherapeutic agent anthracycline. "Anthracycline has been around for years and is one of the most widely used treatments, but it is very toxic, very injurious. We have developed the first test that clearly distinguishes women with breast cancer who will not respond to this treatment," he says.

Other targeted treatments are beginning to be used for cancers that have been highly resistant to treatment, such as lung cancer. Based on test-tube experiments, there was hope that Herceptin might help lung cancer patients receiving chemotherapy. But this January, results of a phase two trial showed Herceptin did not improve these patients’ condition.

However, Iressa, a tyrosine kinase inhibitor developed by AstraZeneca, was approved in 2003 by the FDA for the treatment of lung cancer in patients who have failed both platinum-based and docetaxel chemotherapy, Dr. Bloom says. "Preclinical data had shown that Iressa effectively blocked EGFR signaling in laboratory models and had suggested that it would be even more effective when combined with chemotherapy. The expectation was that it would extend patients’ lives, but the drug didn’t show the same benefit in humans as it did in the animal models. It didn’t show a clinical benefit such as improvement in disease-related symptoms or increased survival. It was approved more as a quality-of-life enhancer."

Expression of the EGF receptor is probably not as important a determinant of who will respond to these therapeutics, Dr. Pegram says. "You have to have it, but alone it’s not sufficient. Perhaps more important is the activation status [of the receptor]."

EGFR PharmDx is similar to the HercepTest because it is an immunoassay and detects HER receptor 1 instead of HER2. "But the similarities end there," Dr. Pegram says, because the HER2 assays offer much better discrimination. "In the case of HER2, whenever it’s overexpressed, it’s constitutively activated. So those tests are much better predictors for Herceptin than this one [is for Erbitux]."

He adds: "All we can do with the current diagnostic is determine whether EGFR is present or not. It’s a good filter and it’s a good first pass for patient selection, but it’s imperfect. The next generation of EGFR assays will hopefully be more specific."

The National Cancer Institute has been seeking to accelerate the development of clinically validated cancer tests, says M. Elizabeth Hammond, MD, a member of a task force that is setting NCI strategy. "The task force has spent extensive time talking about EGFR testing as well as agents active against the receptor and its pathways of activation."

Dr. Hammond, professor of pathology and internal medicine at the University of Utah School of Medicine and director of the Office of Research Affairs at Intermountain Healthcare, Salt Lake City, runs the tissue bank for the NCI-funded Radiation Therapy Oncology Group, which has been involved in trials on head and neck cancer and which is testing agents targeted to EGFR activity.

Erbitux was pending at the FDA longer than expected because there were no good data that the magnitude of receptor activity correlates with the response to drugs, she says. "The reason is there are multiple pathways by which agents active against the receptors or its downstream pathways could be affecting cell behavior—not just one way. Thus, it’s difficult to understand the relationship of the diagnostic test to drug efficacy," Dr. Hammond says.

"Trials now underway are testing EGFR-targeted treatment of head and neck cancers, so we should get information back within three years on that." But the bigger question is whether doing these diagnostic tests will be of any help. "I’m not sure, because patients who don’t have EGFR detected may still respond to drugs. So we might be denying the drug to patients who might respond to it. Only time will tell," she says.

Two types of drugs are active against EGFR, Dr. Hammond explains, and they act by different mechanisms. Erbitux is an antibody targeted against the receptor protein EGFR. Other agents, like Iressa, are inhibitors of tyrosine kinases, which are important components of the cellular activation pathways triggered by the receptor. "Diagnostic tests that measure the protein receptor do not necessarily predict how this protein expression will relate to cellular activation in the tumor being assayed. Thus, the value of the diagnostic test is less specific than the value of tests for HER2 receptor protein, for example."

In all tests used to predict response to a specific therapy, standardization is important so that the right patient population gets the therapeutic drug. "Where EGFR is concerned," says Dr. Hammond, "the threshold for positivity of the test is very low and the instructions for interpretation are very clear. If the standardized test is used according to the manufacturer’s instructions, then it’s very likely the pathology laboratory will be able to interpret the test appropriately and get the right answer. The kit provides all the reagents and the methodology. The drug can be used only in patients who have any EGFR activity, and the diagnostic test can narrow down that number, but the data showing how many people could benefit is all over the map."

In other standardized, semiquantitative tests such as tests for HER2, estrogen receptors, progesterone, and CD20 antibody, "similar kinds of issues arise," she says. "The need for standardization in all of these tests is very high because the response to therapy is directly related to the test result. Pathologists are obligated to carefully understand the test characteristics and interpretation criteria and make sure that the tests are done as accurately as possible."

Dr. Bloom says now that Erbitux and Avastin have been introduced, more targeted cancer therapies will follow. Pathologists will be playing an increasingly central role in these cancer treatments, he says, because they will need to assure oncologists that a target is present or a pathway is activated before the patient is given a particular therapy. "These therapies can range from $10,000 to $30,000 a year. You don’t want to give them unless there is a chance of clinical benefit," he notes.

Dr. Hammond says, "It’s going to be helpful to patients that this test is being offered as part of a standardized kit, so the potential of doing the test correctly is much higher." The CAP is looking into offering a proficiency testing module for EGFR to help pathologists ensure that they and their labs are competent to perform it. "That’s in development right now," Dr. Hammond reports.

Since EGFR PharmDx is a first-generation diagnostic test, Dr. Pegram says, later iterations are likely to have much better predictive ability. NCI senior investigator Janet Dancey, MD, agrees, noting that technologies are rapidly evolving to identify in small quantities of tumor cells the molecular phenotypes determining sensitivity to EGFR inhibition.

"The task of accurately identifying the molecular signature of EGFR inhibitor-induced antitumor activity seems daunting," Dr. Dancey wrote last year in Signal, the journal of EGFR-targeted cancer therapy. But given how quickly biotechnology is progressing, she said, future tests for key biomarkers in cancer treatment could be far more accurate than anything developed to date.

Anne Paxton is a writer in Seattle.