|
|
|
|
CIAO DATE: 12/03
FDA Rules Hinder Cancer Vaccines
On The Issues
December 2002
The Food and Drug Administration must use rigorous drug trials, but these tests should not be so strict as to deny cancer patients access to potentially lifesaving drugs. The FDA must listen to its clinically minded regulators and conduct statistically meaningful tests that still put drugs in the hands of those who need them.
Like most of my cancer patients, Rick didn't want to die without a fight. So he enrolled himself in the clinical trial of a new vaccine designed to prime his immune system so his body could attack his runaway kidney cancer. The trial Rick signed up for was the last leg of testing before the vaccine could be cleared for sale to any cancer patient that needed it. But the vaccine, like Rick, never had a chance.
The focus on cancer vaccines got a boost recently when a promising vaccine for cervical cancer made headlines. The findings, published in the New England Journal of Medicine, said the cervical cancer vaccine developed by Merck worked on 100 percent of the women involved in its first big trial. While it is a significant breakthrough and will save countless lives, in truth, it is not really a cancer vaccine—it is a viral vaccine. It does not target the cancer itself, but the virus that causes it to grow. So it is no help to patients like Rick, whose cancer has already developed.
Still, it has brought renewed attention to the concept of vaccinating someone to fight off cancer and to all the products now in development that could eventually yield a new cure. The biotechnology firm Dendreon is working on a promising cancer vaccine for the treatment of prostate cancer, CancerVax has one for melanoma, and Cell Genesys for lung cancer. There are dozens of similar therapeutic vaccines in development.
The aim of these vaccines is to stimulate the immune system to generate its own cocktails of cells that will attack the cancer naturally. Some try to stop cancer from coming back. In others, the goal is to shrink existing tumors or delay their progression. The problem is that most cancer patients cannot get one of these promising vaccines because the Food and Drug Administration (FDA) is too wedded to its familiar bag of statistical tests and to proving new drugs work by testing them on patients during their last days of life.
Reliance on the Wrong Statistical Tests
The few patients who do get the vaccines do not respond to their unique mechanism because regulatory rules compel companies to test them on patients like Rick, who are already too sick to benefit. The vaccines work by stimulating the immune system to begin fighting off the cancer on its own by generating antibodies that target proteins on tumor surfaces. Since most of the patients receiving these vaccines are already on death's door, their depleted immune systems are hard pressed to respond to the treatments. As a result, the trials are unlikely to yield the statistically favorable results preferred by the FDA, and so the treatments will never be approved.
Yet some of these vaccines have worked very well in patients whose cancer is caught early or is of a less aggressive variety. But the FDA is increasingly requiring companies to conduct randomized and placebo-controlled trials with new cancer treatments, making sure these kinds of patients do not get the vaccines. Patients with treatable cancers do not eagerly enroll themselves in experiments where they might be randomized to receive sugar pills. Companies are hard pressed to recruit enough early stage cancer patients to complete one of these rigorous trials.
The process for testing new cancer drugs has become one big game. Companies test new treatments on patients with terminal cancer because these patients are easier to enroll in clinical trials since they have few attractive alternatives. It is also easier to prove marginal benefits, since these terminally ill patients do not need to be followed as long to prove a new treatment prolonged their lives by a few weeks over those receiving a sugar pill.
Regulatory considerations also alter the types of cancers companies end up targeting. Most cancers are routinely treated with multiple chemotherapy agents. But these drugs also suppress the immune system. So administering them along with a therapeutic vaccine designed to boost the body's immune response is futile.
But that is just what the FDA requires. Rules compel companies to give the vaccines as adjuncts to standard therapy. So companies avoid tumors where standard therapy means chemo, and they will be forced to give the vaccines alongside immune-depleting drugs. That only leaves a few tumors to choose from, such as melanoma and prostate cancer. These are the tumors they go after. Many patients, of course, given the option, would choose vaccines as monotherapy over successive rounds of chemotherapy, where each subsequent round has a substantially smaller chance of curing their disease.
Statisticians vs. Clinicians
The problem inside the FDA stems from a larger struggle pitting strict statisticians, who want drug companies to conduct rigidly structured, large, placebo-controlled trials, against clinically minded regulators more inclined to let common-sense medicine reign over mathematical rigor.
The statisticians absorb themselves in the numerical nuances of drug regulation, emphasizing the need for mathematically rigorous trials to approve new drugs. They favor large, rigorous placebo trials even when it means delaying new drugs designed to treat terminal diseases like cancer. The concern inside the FDA is that the agency will approve a new drug, and confer its imprimatur, even though the new medicine is no better than a sugar pill.
But the statisticians do not need to forfeit their standards to develop clear-cut evidence proving that new drugs work as claimed. In the early 1990s, the FDA allowed companies to test new AIDS medications through a protocol known as the "Treatment IND." Under these rules, new antiviral medications were distributed to AIDS patients through their physicians, without regard to the stage of their disease or other medications they were taking. There was no rigid randomization. No placebos. Doctors kept track of the results as real patients got the medicines just as they would if the drugs were already approved.
The data might not have been as statistically pure as the results the FDA is used to, but so many people ended up getting the new AIDS drugs, that the dirty data that was generated was able to yield statistically meaningful results anyway. The results had the added advantage of replicating how drugs were going to be administered in real-world situations once approved, rather than the contrived confines of a clinical trial. Economists use this kind of raw data every day to reach meaningful conclusions, and so can the FDA.
That would take some creative thinking, which is something confined to a small camp inside the FDA of medical-minded regulators inclined to overlook the statistical shortcomings of some clinical trials when it comes to deadly diseases and drugs with potentially powerful benefits. Drawn from the front lines of patient care, they are quicker to realize the clinical potential of promising new drugs.
The clinicians flexed their muscle during the recent review of another new cancer drug called Iressa, part of a new class of cancer drugs that aim to disrupt the biochemical signals that turn healthy cells into malignant ones.
During Iressa's FDA review, the agency's statisticians cited shortcomings in the drug's clinical trial in recommending that it should not be approved for sale. But the FDA's cancer advisory panel, drawn mostly from the ranks of practicing doctors, disagreed. The panel's chairwoman, a noted cancer doctor told the FDA's actuaries, "I've never seen a lung cancer patient whose cancer went away by itself. Very clearly there are patients whose cancer went away with Iressa."
That spirit is in short supply inside the FDA, where the statisticians generally rule the day, and it is one reason that promising cancer vaccines are stalled in exhaustive trials, many that have little hope of succeeding. The fortitude that would get these treatments into the hands of more cancer patients has not dominated the agency since the height of the AIDS epidemic in the early 1990s, when the FDA put into place a collection of mechanisms that valued medical concerns over statistical rigor in speeding new antiviral medications to the market—a decision credited with saving countless lives.
The FDA briefly embraced these same pathways when it came to cancer, for example, allowing new drugs to be tested on patients without requiring some to get a sugar pill. In most cases, FDA trials need to be randomized and blinded; where half of the patients enrolled in trials receive the new drug and the other half receives a placebo or sugar pill. That way, the results are not tainted by the so-called placebo effect, where patients sometimes feel better because they knew they were receiving a highly-touted new drug.
But the agency has backed away from this relaxation of the rules every since. While tough statistical standards might make sense when it comes to an allergy pill or another Viagra, patients with cancer cannot wait for the FDA to line up all its decimal points.
So long as the statisticians call the shots, a climate where mathematics trumps sound medical judgment will prevail. And patients like Rick who enroll in cancer trials seeking one last shot at a cure, will be victims twice over. First to their spreading cancers, and then to a regulatory culture that favors statistical rigor over sickness, and compels drug companies to test new treatments on the patients least likely to benefit from them.
Scott Gottlieb, M.D., is a resident fellow at AEI.
Email this citation