By Meryl Nass: Twenty years ago, as a newly qualified doctor, I spent a lot of time in the library reading review articles about my patients’ diseases. Twenty years later, my time in the library is too often spent reading about problems and conflicts of interest within the medical establishment.
Here are a few examples: Industry-funded research results in a much higher proportion of studies showing positive results for new drugs, compared to publicly funded research. Flawed regulatory oversight resulted in licensing, then withdrawal, of many dangerous drugs in the past five years. Established protections for human subjects in medical research, which did not prevent the deaths of several subjects in high-profile cases recently, are being undermined.
A recent Journal of the American Medical Association (JAMA) reported that a full nine out of ten doctors on committees that develop clinical guidelines had financial ties to the industry whose products they recommend. Six of ten doctors had financial ties to companies whose drugs were considered in the guidelines they wrote. Pharmaceutical companies paid for the development of 25 percent of the guidelines.
If medical research is being done properly, with good controls and enough subjects for statistical validity, why do so many studies yield answers that are in direct opposition to each other?
It was widely reported recently that mammograms do not save lives; the studies that had claimed they did, were fatally flawed when they were (finally) carefully examined. There is no question that use of mammograms leads to earlier diagnosis of breast cancer than not using mammograms. But this does not result in improved life expectancy. Does it mean that we should stop seeking early detection for breast cancer—that breast cancer is in some way different from all other cancers?
Are we even asking the right question? Is the problem the mammogram or is the problem that aggressive treatment of breast cancer could actually decrease life expectancy in many cases, so that overall there is no treatment benefit in this disease? Is anybody performing solid research to answer the question?
Because there exists a multi-billion dollar establishment that deals with breast cancer in a fairly monolithic way, one is limited as to what questions are allowed to be asked. You can ask, but who will fund your research? Research funded by the federal government is generally constrained to stay within the existing boundaries of disease management, despite its public funding. It will often mirror corporate-sponsored research.
When you don’t ask the right questions in clinical research, you can obtain answers that result in worse patient care. Corporate sponsors of research are not going to spend millions for a trial that could produce an answer in conflict with their goals, if they can avoid it.
Institutional Protections?
Over the last decade there has been a shift backward by regulatory agencies charged with protecting the public health. Although the shift at FDA has been blamed on the 1992 Prescription Drug User Fee charged to the manufacturer to review new drugs and to expedite drug approval, the problems seem to be much more profound. Years ago, a new treatment had to be proven safe before it could be used; during the past eight to ten years, unless there was significant evidence of danger, new drugs were assumed to be safe. In a 1998 survey, FDA’s medical officers reported that standards for drug approval had declined. Last May, Richard Horton, editor of the Lancet, described the FDA’s process for the re-licensing of a drug that had earlier been taken off the market. He explained, step by step, how the FDA had a “two track process, one official and transparent, one unofficial and covert.” The FDA controlled the composition of its advisory committee and its agenda, so the committee would not overturn the agreement already made between senior FDA staff and industry executives. Clearly, there is a big problem at FDA.
A move to weaken human subject protections in clinical research has occurred parallel to this weakening of drug oversight.
The Center for Disease Control (CDC) recently sponsored a trial of post-exposure anthrax vaccine use. The FDA approved the trial. The study’s consent form acknowledged that preliminary data showed anthrax vaccine could cause birth defects. Since, for the preceding two months, antibiotic treatment had been 100 percent successful at preventing anthrax in those exposed, it was not at all obvious that vaccination offered any additional benefit. Yet pregnant women were invited to enroll as subjects.
But that wasn’t the end of it. The FDA approved the license for anthrax vaccine and approved a new anthrax vaccine label, which became public five weeks after the CDC study began. The new label clearly states that no animal experiments have ever been performed to determine the vaccine’s effect on pregnancy.
What logic led both the CDC and the FDA to experiment on human fetuses in the complete absence of animal fertility data? These agencies have lost sight of their mandate to protect public health. Their lack of ethics might have been influenced by the Defense Department’s contribution to their budgets, which amounted to $2.5 billion last year for the CDC.
Additional moves are afoot to weaken the protections for children in clinical research. The Jesse Gelsinger case, in which a teenager died from participation in a gene therapy experiment from which no personal benefit was expected, demonstrated that fully informed consent is often missing in clinical trials. Informed consent is presented to potential participants by the researcher, who has a vested interest in signing up subjects. Its oversight by institutional review boards tends to be cursory. In the Gelsinger case, the principal investigator, along with the University of Pennsylvania, had a large financial stake in the outcome of the experiment.
Perhaps half the drugs used in children have never been licensed for pediatric use. They are prescribed “off-label” by clinicians. For many drugs, such as antibiotics, which have demonstrated safety and efficacy over many millions of doses, this is not a problem. Both patients and doctors are perfectly satisfied using such drugs in the pediatric population, on or off label.
But in the case of other drugs, such as psychotropic medications, many doctors are loathe to prescribe for children without adequate pediatric testing. Drugs that are given indefinitely are better moneymakers than antibiotics, which are only used for ten days at a time. So expanding approvals for chronic drug use into the pediatric age group could yield handsome rewards.
Perhaps as a result, the rules for using children in clinical research are being undermined. No longer would a child need to clearly demonstrate potential benefit from a new treatment before being enrolled in a trial; proposed rules would allow a child who is “at risk of” the condition to be used as a subject. But most of us are “at risk of” most diseases.
Furthermore, new consent forms have been developed that allow adolescents to provide a modicum of “informed consent.” (They were used in CDC’s recent anthrax vaccine trial.)
Using Bioterrorism Fears
You cannot compel people to become experimental subjects: that is the legacy of Nuremberg. If a drug or vaccine is not fully licensed, it cannot be forced on anyone.
But it is desirable to have drugs ready to counteract a chemical or biological attack. If the illnesses anticipated from an attack do not occur naturally in the population, one cannot test the new drugs for effectiveness. It would be unethical to expose people to a chemical or disease just to test whether the new drug is truly protective.
That results in a conundrum: if you have to prove effectiveness in human trials to license a drug, but you can’t do the trial, then you can’t license the drug. If the drug is not licensed, it can be used with informed consent, but you cannot force people to take it. The Defense Department was not satisfied with that. There is no provision for informed consent on the battlefield and a soldier who refuses an experimental treatment could endanger the lives of his colleagues, so they said.
President Clinton issued Presidential Order 13139 to deal with this situation. It allowed the president, in consultation with the Secretary of Defense and the FDA Commissioner, to require that troops take experimental drugs or vaccines in special circumstances.
That should have been adequate to take care of the situation, but the federal authorities were still not satisfied. The National Research Council (of the National Academies of Science) was contracted to report on protecting troops from bioterroism. “How can we ensure safety of troops if we have to go through an onerous two or five years of certification [for new drug approval]?” asked Robert Love, the study director. His June 2001 report recommended that the Army seek exemptions from some regulatory approval processes to speed up the development of new medical treatments.
CDC did its part by contracting with Laurence Gostin, a law professor at Georgetown University and professor of public health at Johns Hopkins University, to create a Model Law that the states would be encouraged to use. It would give state officials the authority to involuntarily quarantine and vaccinate citizens, among other things. Does it seem odd that Health and Human Services is giving the states a blueprint to consolidate control over their citizens in the event of bioattack?
The FDA had already embraced the new regulatory culture. How better to both speed up drug approvals and save the president the political cost of contravening the Nuremberg Convention than by weakening the current requirements for drug licensing?
After acknowledging that the effectiveness of bioterrorism drugs could not be tested in humans, and therefore animal efficacy tests should be used instead, FDA prepared to throw out the baby with the bathwater. In 1999, hints that human safety testing would be jettisoned as a requirement for licensure began to appear.
The 1999 Annual Report for FDA’s Center for Biologics said: “A research program to produce vaccines, therapeutics and drugs to treat [bioterrorism] outbreaks faces the challenge of not being able to proceed with Phase III efficacy clinical trials. Given ethical and safety concerns that would rule out infecting human subjects with a deadly organism in order to test a vaccine or therapeutic for efficacy, trials with humans cannot be undertaken. Therefore, the regulatory process for approval of treatments must be modified to permit the emergency use of antibiotics, therapeutics and vaccines that have been shown to be safe and effective in animal models.”
Wait just one minute. Nobody needs to be infected with anything to test drug or vaccine safety. All you do is administer the drug or vaccine and watch the recipient for possible adverse effects. Could this leap of faith in animal models have been a mistake? After all, it is universally acknowledged that human adverse effects cannot be extrapolated from animal tests. Each species responds uniquely to a drug or vaccine. Vaccines that are safe in some species can be fatal in other species and this cannot be accurately predicted ahead of time.
Unfortunately, it looks like there is no mistake. The director of FDA’s Center for Biologics, Kathryn Zoon, published a paper in “Emerging Infectious Diseases” in which she reiterated the call to fully license drugs for bioterrorism in the absence of any human testing. She said that once licensed, the safety of the drugs can be assessed. It is hard to reconcile this philosophy with Zoon’s role as a federal regulator charged with protecting the public health.
The latest episode in this saga concerns Congress’s role in bioterrorism prevention. A bill designed to fund federal bioterrorism efforts, called the Public Health Security and Bioterrorism Response Act of 2001 was passed in December by both the House (HR 3448) and the Senate (Senate Amendment 2692) on the same day it was introduced.Both the House and Senate versions of the bill contain a provision mandating that FDA finalize and implement a 1999 Notice of Proposed Rulemaking. This action would allow animal efficacy tests to be sufficient to fully license drugs and vaccines intended for bioterrorism. Although safety testing is not explicitly addressed in the bill, given the statements made by FDA above, it appears that safety testing in humans may be waived as well, as a requirement for licensure.
Another way to look at this bill’s provision is as a way of getting around the absolute requirement for informed consent. As was pointed out earlier, people are allowed to use experimental drugs and vaccines, as long as the FDA has approved the experimental use and the patient or subject has provided informed consent. By licensing what would previously have remained an experimental drug, one opens up the possibility of forced use, with no need for informed consent. The Nuremberg Convention, first nibbled at by mandating use of experimental drugs in the Gulf War, looks like it is about to be completely overturned.
This bill is now in conference committee. Congress should be informed that despite having undergone human safety and efficacy tests to be licensed, many drugs and vaccines have still had to be withdrawn from use due to severe side effects, including death. These side effects were usually not discovered until the drugs were given to large numbers of people.
In the case of drugs and vaccines for bioterrorism, it is likely that, following an attack, the entire country will receive a drug or vaccine in a crash program, over days or weeks. This leaves no time to assess the adverse event profile of the drug in smaller numbers of people, before it is given to the entire population. If the Model Law is used, forced acceptance of drugs or vaccines that have never been tested in a single human could be demanded of the entire U.S. population. One poor choice of a drug or vaccine that is later found to be dangerous could have a dire effect on a very large number of people.
Human safety testing is not something we should allow to go by the board.