How Well Tested Are New Cancer Drugs?

How Well Tested Are New Cancer Drugs?


by Maryann Napoli

With a certain amount of regularity a new cancer drug makes headlines, generating an enormous amount of hope as well as pressure to make the product swiftly available. In time, we usually learn that the drug’s benefit is much more modest than was originally portrayed in the media, and soon oncologists begin to prescribe the drug for other forms of cancer without waiting for clinical trials to prove its effectiveness.

Tumor Shrinkage

Contrary to popular belief, drug companies are not required to prove that their drugs prolong survival. Until the mid-1980s, all cancer drugs were approved solely on the basis of what researchers call the “tumor response.” In other words, a drug company needed only show that the drug caused a tumor to shrink, not necessarily to disappear.

Years ago, a change in the approval process was recommended by the FDA’s own Oncologic Drug Advisory Committee. The committee members, primarily cancer experts unaffiliated with any government agency, knew that tumor shrinkage often has little or no relation to survival. The committee proposed the novel idea that a drug company should be required to prove that a drug provided some benefit that was meaningful to the patient, such as increased survival or an improvement in symptoms. The committee argued further that the potential benefit of tumor shrinkage did not necessarily outweigh the substantial toxicity of cancer drugs.

This recommendation was made in the mid-1980s, but change at the FDA comes slowly, as a recent assessment of new drug approvals has demonstrated. From 1990 through 2001, the FDA approved 66 new cancer drugs. Prolonged survival was not proven for 48 drugs. And tumor response was the basis of approval for 35 drugs.

Variations in End Points

” The FDA has a certain amount of regulatory flexibility to make an assessment of the side effects versus the efficacy of new products,” said Richard Pazdur in a telephone interview. The director of oncology drug products at the FDA’s Center for Drug Evaluation and Treatment, Pazdur had been asked why so few new drugs have been proved to prolong survival.

“The drug company must prove that its product provides a longer life, a better life, or a favorable effect on an established end point for a better life,” he explained. The FDA’s flexibility comes into play on the last item, “an established end point for a better life.” One example of an established end point, says Pazdur, is a complete response in leukemia: that is, the bone marrow is normal and the blood counts have normalized.

Even so, there is a hierarchy of established end points. “Survival is the gold standard of end points because it is the most meaningful to all people,” Pazdur said. There can be no misinterpretation when the end point concerns survival: the patient is either dead or alive. “Whereas the other end points, such as tumor response rate, are usually determined by X-rays or scans,” he continued. “There can be variations in the radiologists, the techniques of how the X-rays or scans are obtained, which can make it confusing and unreliable. Also, there is a subjective judgment in reading these X-rays and scans.”

Accelerated Approval

These cautions notwithstanding, the FDA still allows the use of tumor shrinkage as the sole end point in the approval of certain drugs. In 1992, the agency introduced an accelerated approval (AA) process. The idea behind AA is to get drugs quickly to advanced-cancer patients in whom all available options have failed. Consequently, tumor shrinkage was the sole basis of the AA for 10 of 11 new drugs. The rationale: Shrinking a lung tumor might, in the FDA’s view, be “reasonably likely” to alleviate breathing difficulties.

The testing for AA is minimal. The new drug is given to about 30 or so participants who have run out of options. In what is called a phase II trial, there is no comparison group: everyone in the study gets the new drug. Consequently, this type of trial is not likely to provide a true picture of the drug’s toxicity or efficacy. That’s why a drug given AA must continue to be studied to see if it provides any benefit in terms of increased survival or symptom improvement. “The drug companies usually do a large trial in which the new drug is compared to the standard drug,” said Pazdur. “This usually takes two to four years.” And what if the drug company does not comply? “We have a process for rapidly removing the drug from the market,” Pazdur replied.

Still, some not so well tested drugs are available for several years following an AA, and oncologists are free to prescribe them for cancers other than the type for which the products received approval. Or, more often, oncologists can add the new product to a multiple-drug regimen that in itself has never been studied. “Yes, this is called off-label use, which is fairly common in the practice of oncology in the U.S.,” Pazdur agreed. “But this involves the practice of medicine, and the FDA does not control the practice of medicine.”

Unless their oncologists tell them, cancer patients do not know whether they are being given a drug off-label. “We would like patients to read the drug label to see the approved indications and contraindications,” advised Pazdur, who said that the information is freely available at the FDA’s web site.

For More Information

To read a drug label, go to or to the Physicians’ Desk Reference (PDR), which is updated yearly and available at most local libraries and large bookstores. The label is daunting for its extensive length, tiny type and medical jargon. It is, however, the only source for results of the FDA-required tests. The section entitled “Indications” will identify the purpose(s) for which the drug has been proven beneficial. If you can’t find your type and stage of cancer listed under “Indications,” this signals off-label use.

Unfortunately, only the rare cancer patient well versed in clinical trials will be able to discern whether the label is identifying a drug that has gone through the less-rigorous AA process. For example, capecitabine was given AA (on the basis of tumor response) for advanced breast cancer in 1998. The 2002 Physicians’ Desk Reference describes capecitabine’s testing as a “phase II, single-arm trial,” which signals AA. Another clue can be found under “Indications,” where response rate is described as the basis for the drug’s use for metastasized breast cancer. However, the much watered-down patient version of the capecitabine label (which appears after the information aimed at professionals) does not include an explanation of these terms. The patient’s label makes no mention of the fact that the drug was approved through the accelerated process. Capecitabine is sold under the brand name Xeloda.

Go to the web site of the National Cancer Institute, and then click into the following succession of options: “treatment,” “chemotherapy,” and “newly approved cancer treatments.” You will find explanations of phase II trials, off-label drug use (complete with Q and A: “Can off-label drug use be harmful?”), and many other relevant terms. People without access to the Internet can call the National Cancer Institute toll-free at 800/4-CANCER to get this information mailed to them.

When recommending a new cancer drug to a patient, oncologists will often quote “response rate” without explaining what it means. Ask. This article has addressed how new drugs receive FDA approval. But once on the market, they are often studied eventually in large randomized, controlled trials as part of a multiple-drug regimen.

Ask for the evidence to support a proposed chemotherapy regimen. Here’s one way to phrase the question: “Can you give me a citation for the studies that show this drug or drug-combination will benefit people with my stage and type of cancer?”

The citation will allow you to do a Medline search (ask the librarian at your local public library) to locate the study and determine the exact nature of the benefit. Medline is a service available through the National Library of Medicine. It provides free access to the abstracts, or summaries, of the studies published in a large proportion of the world’s medical journals. Some public libraries will retrieve the entire article at no charge, but the article can also be purchased on-line.

Maryann Napoli is the associate director of the Center for Medical Consumers in New York City. This article is adapted with permission from HealthFacts, the organization’s monthly newsletter.

Join Thousands of People & Receive - Advanced Health & Wellness Monthly Newsletter
Join Our Wellness Newsletter!