Editor’s Note: This story, first published in 2005, has been updated.
At first, Nancy J. Brown, M.D., wasn’t certain whether or not she was seeing a pattern among her patients.
It was 1992, and the young clinical pharmacologist at Vanderbilt University Medical Center noticed that some of her patients seemed to have an allergic-type reaction to ACE-inhibitors, medications used to treat hypertension and reduce the risk of heart disease and kidney failure.
Not commonly, but often enough to notice, some of them were coming in with symptoms of angioedema—swelling of the lips, face, tongue and throat. Even more disconcerting, it struck her that an inordinate number of patients with side effects were African-American.
Brown decided to pursue her hunch. First, she called various drug companies, but few had readily available data on adverse reactions. Next she spoke with Marie R. Griffin, M.D., MPH, a Vanderbilt pharmacoepidemiologist, who contacted colleagues at the U.S. Food and Drug Administration to see if there had been any similar reports of adverse events.
As it turned out, a company had just submitted an application to the FDA for a new ACE-inhibitor drug, and had reported seeing higher rates of angioedema than anticipated. That company had included more African-Americans in its trials than previous studies had done.
Like sleuths, Griffin and Brown then examined the Tennessee Medicaid database, comparing rates of angioedema among both black and white patients on ACE-inhibitors with patients on calcium channel blockers, a different class of medication for reducing blood pressure.
“When we controlled for (type of medication) we found that blacks (on ACE-inhibitors) were four-and-a-half times more likely to have angioedema than whites,” says Brown, now a professor of Medicine and Pharmacology. “The other thing we noticed was that there were a number of people who’d had episodes of angioedema who’d been left on their ACE-inhibitor.”
In other words, even when patients presented with angioedema, many physicians hadn’t made the connection that the medication was the underlying cause.
Part of the reason was because angioedema, even among blacks, is a rare event. But also, while some people took only a pill or two and quickly experienced dramatic swelling of their tongue or lips, other patients had been taking the medicine for months or even years before the side effect kicked in. Still others never experienced bad reactions.
Brown, who has earned national recognition for her research on blood pressure regulation, now has a grant to try to weed out the biologic mechanisms for angioedema induced by ACE-inhibitors. One of her concerns, she says, is, “How do you study something that is still a pretty rare side effect?”
At first glance, it would seem that if a drug has gone through the rigors of discovery and development, through clinical trials and FDA approval, any adverse side effects should have already been observed and noted. After all, isn’t it the responsibility of the FDA to ensure that only safe drugs are put on the market?
The answer is yes—if only it were that easy.
Deadly elixir
Of late, scientists and regulators have been anxiously grappling with the problems posed by post-marketing surveillance of approved pharmaceuticals. What may seem like a miracle cure in controlled clinical trials, several thousand people strong, may become a totally different animal once it is released to the open market and prescribed by tens of thousands of physicians to millions of patients.
Ignoring strict warning labels, many patients take medicine off-label, meaning they take it for illnesses for which it hasn’t been approved. They also take combinations of drugs that counteract each other’s beneficial effects, under-dose to save money or take more than the recommended dose to try to get quicker relief.
Even more confounding is the fact that two patients with the exact same disease can follow the exact same treatment plan—and the drug works beautifully for one, but doesn’t help the other patient at all.
All of these challenges have served to usher in a new era of pharmacoepidemiology—the study of pharmaceuticals among populations of patients. According to Wayne A. Ray, Ph.D., director of Division of Pharmacoepidemiology at Vanderbilt, the field first emerged in the aftermath of two therapeutic disasters.
In 1937, a Tennessee drug company began selling Elixir Sulfanilamide, a sulfa drug used to treat streptococcal infections. More than 100 people, many of them children, died after taking the medication, which was dissolved in a toxic component of antifreeze. In response, Congress passed the Pure Food, Drugs and Cosmetics Act of 1938, requiring products to be rudimentarily tested for safety before they reached consumers.
The second disaster involved the drug thalidomide, marketed in the 1950s and early ’60s as a sleeping pill and a cure for morning sickness in pregnant women. Thalidomide caused profound birth defects in more than 10,000 children in 46 countries before it was pulled from the worldwide market.
Although a pharmacologist at the FDA, Frances Oldham Kelsey, M.D., Ph.D., prevented thalidomide from being marketed in the United States, the side effects were so catastrophic that in 1962 the United States began requiring that drugs undergo well-controlled studies demonstrating efficacy before they can be sold to consumers.
The problem in our present day and age, says Ray, is that the rather cumbersome guidelines for ensuring safety and efficacy of drugs slowed down the process of getting drugs to market, and in the 1980s consumers and pharmaceutical companies alike began pushing for quicker approval.
The upshot, he says, is that “… to get drugs to market in a reasonable amount of time, we put them through phase I, II and III trials, but we still don’t know much about them. We know enough about a drug so that people can begin using it, but not enough about its long-term effects. So we have to keep studying it.”
Unfortunately, that is where the existing system falls short. “Following a drug once it’s on the market is a free-for-all,” he says.
Reporting the signal
The system the FDA uses for conducting post-marketing surveillance of drugs, MedWatch, relies on doctors, nurses and pharmacists to report safety concerns, based on suspected serious adverse reactions they have observed. There are two instances in which this type of reporting system works well—if a problem arises that is extreme or unusual, or if a problem appears within a large proportion of consumers.
In the first instance, a drug may produce such an unusual, distinctive form of toxicity that physicians immediately take notice, such as the idiosyncratic birth defects among pregnant users of thalidomide.
Another example is the drug-induced arrhythmias discovered in consumers of the antihistamine Seldane.
“Seldane was given to millions of patients before the potential for severe drug reactions causing life-threatening arrhythmias was recognized,” says Dan M. Roden, M.D., who directs the John A. Oates Institute for Experimental Therapeutics at Vanderbilt. “The adverse effects only occurred in patients who were using Seldane in combination with other drugs.”
Although the arrhythmias caused by Seldane toxicity were so rare that they were hard to detect, their uniquely abnormal patterns enabled physicians to eventually track similar peculiar rhythms among patients and trace them back to use of the drug. Seldane was withdrawn from the market in 1997.
A much more difficult problem to identify, says Roden, is drug toxicity that leads to an increase in a common event, such as heart attack or stroke among the elderly. The situation becomes even more complex if patients have been taking a medication for weeks, months or years before they suddenly have adverse reactions. Physicians may have no reason to suspect the drug.
“Suppose a drug increases colon cancer by 20 percent. Unless they do a prospective clinical trail, we will never know,” Roden says. “My personal bias is that these kinds of unrecognized drug actions are all around us.”
A key to determining both the benefits and detriments of a particular drug is finding the first “signal”– the indication of some unanticipated side effect—and reporting that signal to the FDA.
Many epidemiologists and pharmacologists in the field claim that this mode of communication is inherently flawed. “It’s hit or miss,” admits Vanderbilt’s Griffin, professor of Preventive Medicine.
Currently, MedWatch has little teeth for affecting change even if officials receive information warranting further investigation of adverse events. They can demand that a company conduct post-marketing surveillance on drugs, but they have no power to force a company to do so.
In effect, the FDA has only two options: demand a change of label, or to withdraw the drug from the market.
“The agency (FDA) has only a nuclear bomb (withdrawal) that they can bring to bear…or a powder puff, the powder puff being a change of label, which nobody reads and is totally ineffective,” says Raymond L. Woosley, M.D., Ph.D., a former Vanderbilt clinical pharmacologist whose work on Seldane led to the discovery of a safer antihistamine now marketed as Allegra.
Beyond that, even if a drug is found to cause a dangerous side effect in rare cases, there is no standardized mechanism for determining what happens next or what should be deemed “acceptable risk.”
“Should we deprive 100 people of ACE-inhibitors to prevent one episode of angioedema?” asks Griffin. “That’s a difficult question.”
Search for safety
In the past, drugs were usually given in limited courses of days or weeks to address an acute medical problem. Many of today’s medicines are administered to enhance quality of life and to prevent disease from occurring, and can be taken for years on end.
“When you’re giving drugs to generally healthy people, the standards have to be higher,” she continues. “If a drug is not saving your life, it better not be killing you.”
Seeking to advance the “optimal use” of drugs, medical devices and biological products, the federal government in 1999 launched the Centers for Education and Research on Therapeutics (CERTs) initiative.
Seven centers, including one at Vanderbilt, conduct a wide range of research on drug use and safety. Their reach is limited, however.
“When you try to take research into practice there’s a lot that’s lost in translation,” explains Ray, who heads up the Vanderbilt CERT. “We address drug safety, but also the problem of clinicians using interventions appropriately. For example, it doesn’t help if beta blockers prevent deaths if nobody prescribes them.”
Meanwhile, aggressive marketing, including direct-to-consumer advertising, has created a demand for new drugs, even among patients for whom the compounds may not be effective and may even be dangerous.
The ticket to better post-marketing surveillance might have to come from a complete redesign of the FDA’s framework.
Currently, all the years it takes a drug to go through clinical trials and approval counts against its 20-year patent life. When the patent expires, competitive generic compounds can join the pool.
Former Vanderbilt clinical pharmacologist Alastair J.J. Wood, M.B., Ch.B., has proposed that the FDA reward pharmaceutical companies that conduct post-marketing studies of their products by lengthening their period of exclusivity—postponing the right of generic knockoffs to compete against them.
Wood, Woosley and others also have called for the establishment of an independent agency to monitor drugs once they hit the market.
In 2005, in response to ballyhoo by physicians, legislators and consumer watch groups, the Department of Health and Human Services announced that the FDA would indeed create a new independent Drug Safety Oversight Board.
Its purpose is to promote a culture of openness and to provide emerging information to patients and doctors about the risks and benefits of medications. The board is composed of FDA employees and various government employees who consult with medical experts and patient and consumer groups.
While this was a step in the right direction, some experts wondered whether it was sufficiently independent or had the authority and resources necessary to do the job.
“Society takes a calculated risk every time we release a new drug onto the market,” Ray notes. “There needs to be someone, some independent authority, in place to make hard decisions.”