AMSA Convention 2016 Logo

Up for Review

The New Physician May-June 2004
Hundreds of drugs enter clinical trials each year, but only about one-fifth will eventually win FDA approval. How does the process work? Let’s take a look….

The line at the pharmacist’s counter is a constant flow of coughs, colds and chronic diseases, all clutching little squares of paper with unintelligible writing. As their physician, you might have written it out just 10 minutes before—a defense for the sick and the hurting. Behind the counter, the pills and potions offer a dizzying array of help. And for the price of their co-pays—or, gulp, full price for those without drug coverage—the pharmacist doles out the medicines they need, when they need them, even 24-hours-a-day in many cases.

The agency that decides which of those bottles and pill packs may sit on the pharmacist’s shelf and how they may be used also sees a constant flow at its door, but the turnaround time for drug companies seeking approval from the U.S. Food and Drug Administration (FDA) is much longer than the seconds it took for you to scribble the prescription—about a year longer. In fact, the time the FDA takes and method it uses to approve drugs have been contentious issues among both Congress and watchdog groups for decades. And even the fix lawmakers and industry leaders came up with in 1992 generates complaints. So what takes so long for a drug to get from the chemist’s lab to pharmacists’ shelves? The answer to that question begins with a look at the process itself.


Congress created the FDA in 1906, but it wasn’t until President Franklin Roosevelt signed the Food, Drug and Cosmetic Act in 1938 that lawmakers gave the agency the power to approve drugs for sale based on newly required safety tests. And for about 25 years, safety was all a drug had to prove. Then the 1962 Kefauver-Harris amendments mandated drug companies show their products are safe and effective before marketing them, which established the risk–benefit scale on which the FDA still weighs each chemical compound brought before it.

The safety tests, conducted and funded by the pharmaceutical company or government agency developing the drug, involve a series of steps triggered after positive feedback from preclinical research on animals and an institutional review board. Before a drug can move into the clinical study phase, the FDA’s Center for Drug Evaluation and Research (CDER) encourages a meeting between the agency and the applicant. At the least, the company must file an investigational new drug application notifying the FDA of its intent to test the drug on humans and seeking permission to ship the unapproved drug across state borders to clinical investigators in other states.

If all goes well, the drug company can initiate Phases I through III clinical trials, in which humans take the potential drug for the first time. Phase I looks at the drug’s reaction on healthy volunteers. Phase II helps determine the drug’s effectiveness and any potential side effects. And, after another meeting with the FDA to determine whether to continue testing, Phase III expands the number of subjects—sometimes to thousands of volunteers—to generate more data on the benefits and risks.

It’s a lengthy process—often taking years of study—and a huge and costly gamble to the drug company, according to a study released last year by the Tufts Center for the Study of Drug Development (CSDD). The average cost of developing a drug, from the first ray of hope in the lab to the new bottle behind the pharmacist’s counter, is $897 million, and only 21.5 percent of the hundreds of drugs that enter human trials will eventually win FDA approval.

The FDA uses the data from the preclinical and clinical phases to determine if a company is allowed to put a drug on the market. All of this information—pages and pages of binders—gets submitted to the FDA along with samples of the product labeling in a new drug application. The documentation required in the application is supposed to tell the whole story, including what happened during the animal and clinical tests, what components the drug consists of, how it behaves in the body, and how it will be manufactured, processed and packaged, especially noting any quality controls, according to the FDA Web site. (The agency declined repeated requests to comment for this article, referring The New Physician to the FDA Web site.)

The breadth of materials that accompany a new drug application makes for some extensive reading for a review team consisting of a physician, a chemist, a microbiologist, a pharmacokineticist, a pharmacologist and a statistician. According to the FDA, their job is to answer the question, “Does this drug work for the proposed use?” while looking for adverse effects, weighing the benefit-to-risk relationship.

Sixteen discipline-specific advisory committees support the review teams, another product of the Kefauver-Harris amendments. Advisory committees consider applications whenever reviewers require assistance on a specific issue, and while the politically appointed experts never have the final say in the approval of a drug, reviewers use them to confirm or redirect the team’s thinking, says pharmacokineticist Funmi Ajayi, Ph.D., who first served on and then led a review team at the FDA between 1991 and 2002.

For example, when the emergency contraceptive pill Plan B went before the FDA in an attempt to gain over-the-counter status, review teams sought comments from both the Nonprescription Drugs Advisory Committee and the Advisory Committee for Reproductive Health Drugs, which heard presentations in December about Plan B’s safety, effectiveness and the risks for eliminating physicians as the link between patients and the drug.

Sometimes reviewers need all the help they can get. “It was not unusual for some of these applications to come to the FDA via an 18-wheeler that backs up to the loading dock and unloads skid after skid of documents,” says CSDD director Kenneth Kaitin, Ph.D.

And while the advent of electronic files has cut down on that somewhat, Ajayi says the trucks were still showing up while she was at the FDA. “The workload is always high, and I believe it will always be high,” she says of her old job. Medical reviewers averaged 10-hour days and regularly put in time at home on the weekends. “We joked that it was a never-ending story.” As a result, turnover was and continues to be high, she says, and her 10-year stint was unusual.

This immense workload stalled approval time and is part of the reason why for many years the FDA heard complaints from industry officials and members of Congress about the long approval process. Prior to 1992, average approval times for new drugs was between 30 months and three years, according to Kaitin. Today, companies wait about a year for the FDA’s OK.

So what changed? Meet PDUFA.


Although passed under FDA Commissioner Dr. David Kessler, the Prescription Drug User Fee Act of 1992 (PDUFA) was the brainchild of President Reagan’s FDA Commissioner Dr. Frank Young—his attempt to combat the growing responsibilities and stagnant funding Congress saddled the FDA with during the 1980s, according to New York Times reporter Philip Hilts’ recent book, Protecting America’s Health: The FDA, Business, and One Hundred Years of Regulation.

The law is designed to provide additional money for the agency by having pharmaceutical companies pay for each new drug application they submit. The fees are set yearly—for fiscal year 2004 each application that requires clinical data costs $573,750 and $286,750 for those not requiring clinical work—and this year the agency expects to collect $250 million from the program—a substantial chunk of its $1.7 billion budget.

There are some exceptions: Applications for orphan drugs—those to combat rare diseases that will have little mass-market appeal—sometimes have fees waived, as can applications from less well-heeled companies. “[The fee is] not a lot to a Merck or a Pfizer, but to a lot of the smaller companies it is a tremendous burden,” Kaitin says.

In return for the fee, pharmaceutical companies have a stopwatch put on the their application reviews: 10 months for standard applications and six months for those submitted under the FDA’s priority drugs program, which deals with such high-profile diseases as AIDS and cancer. Kaitin cautions these deadlines, however, are expectations of when action will occur, not total approval times. Under PDUFA, standard applications take an average of 12 months for approval, although the agency tries to keep high-priority drugs to about six months.

Thanks to this give-and-take, PDUFA has been wildly popular among industry leaders. The Pharmaceutical Research and Manufacturers of America, which declined to comment for this article, put out a statement two years ago calling PDUFA “an enormous success for patients,” citing reduced review times and the 712 drugs the FDA approved since its passage, compared to the 473 in the previous decade.

The program was able to cut review times so drastically because PDUFA funds are required to remain with the divisions at the FDA that handle drug and medical device approvals. Most of the money pays for increased staff—the agency has already added the equivalent of 1,201 new full-time positions since 1992, the majority of which have been within CDER, and expects to increase that by another 293 by the end of fiscal year 2007—but some also goes toward operating costs.

While PDUFA has given the agency more hands on deck, it has also given the industry a window into the decision-making process, says Dr. Nora Zorich, the medical director of the pharmaceutical division at Procter & Gamble, who has shepherded several drugs and the food additive Olestra through the application process, as well as overseen the switch of the heartburn medication Prilosec from prescription to over-the-counter. Prior to PDUFA, “as a sponsor, you had no idea what was going on,” she says. “What PDUFA did was kept those people on your project more than less. I know what’s being submitted; I have committed to a 12-month clock. I have a chance to plan for that.” And so does the agency. “They know four months down the road, they might need some help from pharmacology…. There’s something to be said for planning for extra help when you’re going to need it. PDUFA encouraged the agency to bring that kind of discipline into their work.”

Under the law, CDER staff must now respond to industry requests for meetings to hash out problems and update progress within 30 to 75 days of the request, depending on the type of meeting, Kaitin says.

And the meetings are important. Prior to 1992, many companies didn’t contact the FDA at all during the drug development process, Ajayi says. “Being a reviewer, you were dumped with information that had been gathered over three, four or five years…. You’re at the end of the process, and it was difficult to say, ‘You know, we could have gotten more information if you designed your study [differently].’ You don’t want them to have to go back. It’s like crying over spilt milk.”

Former Commissioner Dr. Mark McClellan, who left in March to head the Centers for Medicare and Medicaid Services (see “No One to Steer the FDA,” p. 22), told a group of industry officials in November that the process is designed to make research and development less costly. “A big contributor of those costs is in the increasing cost of clinical trials,” he said. Efficient trial design will provide the most information in the least amount of time, he added. “We’re trying to make the process more predictable.”

So now, a company can call a meeting with reviewers—and get it within 75 days—before establishing study parameters to ensure the results will be data the reviewers will need to make their final decision. “What the user-fee system has allowed is the ability to have more hands and more brains for thinking. There is more hand holding, more collegial relationships,” Ajayi says. “One enjoys the type of relationship we have with the pharmaceutical industry. It is much more a partnership.”

It’s a change Zorich says she sensed particularly in 2000 when Procter & Gamble submitted the osteoporosis drug Actonel for approval. “I thought the agency very much understood the new clock,” she says. “It’s within their purview to ask for more data…. I would agree that the agency is pretty rigorous, but as a consumer, I have to say, would I want them to be any other way?”


Not everyone is as satisfied with the approval process. Rep. Henry Waxman (D-Calif.) wrote the original PDUFA bill and has supported its two successful reauthorizations, but as a longtime critic of the FDA, he still has some concerns. “Now, some of the problems can be blamed on the agency being underfunded for decades. User fees solve some problems, but they created others as well,” he says.

Having the industry the FDA regulates contribute so heavily to its budget creates a conflict of interest, contends the watchdog consumer group Public Citizen. But pharmacist Larry Sasich, a research analyst at its Health Research Group, says the bigger problem with PDUFA is that all the extra help to approve more drugs more quickly has created a backlog of work for FDA staffers who focus on post-approval regulation—ensuring the newly approved drugs don’t start causing side effects that didn’t turn up in the development studies and that drug companies market them as they said they would. “All these new drugs required extra monitoring for advertising, etc., and none of that [PDUFA] money was available,” he says.

“The last time the legislation came before Congress, I fought that some of this money could be used for post-marketing study,” Waxman says. He won that battle and continues to be a fly in the ointment over FDA funding. “There’s a lot they have to do, and they don’t have the resources to do so.” (For more information on drug companies’ off-label marketing, see “Not FDA-Approved,” p. 12.) “The other side of it is when we wrote the user-fee legislation, we required the FDA to approve drugs more quickly,” Waxman says, adding that sometimes a faster process leads to more mistakes.

And mistakes have happened. The period between September 1997 and June 1998 was particularly bad for the FDA. During those 10 months, five drugs had their FDA approvals withdrawn for safety reasons—often because they were killing patients—and editorials across the nation called into question the hastened approvals. Hilts calls 1997 a “nightmare year,” reporting that five drugs approved that year were also withdrawn in later years, and two more drugs received black-box warnings on labeling—the strongest advisory the FDA can insist on while still allowing a drug on the market. He hastens to note that 1997 was not a normal year in that a huge glut of drugs was submitted for review. And Ajayi points out that drugs submitted in the mid-1990s were under development before PDUFA was passed and so were created without the FDA’s extra guidance.

But Sasich thinks differently. “It’s a pressure to approve. The FDA has always been between a rock and a hard place,” he says. “The problem lies with Congress in that there hasn’t been real congressional oversight since about the mid-’80s. Previously, if the drug gets recalled, Congress would look into it. [But] in terms of drugs that have come off the market, it’s the equivalent of 15 plane crashes with the [National Transportation Safety Board] saying, ‘Too bad.’ In a democracy we should not have to do that.”

Sasich cites a 2002 U.S. General Accounting Office report that found the percent of drug withdrawals at the FDA went from 1.96 percent between the pre-PDUFA years 1989 and 1992 to 5.34 percent between 1997 and 2000. However, the report notes that factors including physician and patient error contributed to some recalls, and that it only takes a couple of withdrawals to spike the percentage.

But Kaitin adamantly refutes claims that PDUFA has resulted in more drug withdrawals. “This is just not the case. FDA documents [that] the number of removed drugs has remained the same pre- and post-PDUFA.”


The turn-of-the-21st-century withdrawals, coupled with a series of damaging court decisions against the FDA, called into question the power of its regulatory authority. Patient advocates have been saying since 1992 that the cozying up of the FDA to the industry it regulates is troubling, but Ajayi says no one gets a free pass. “The ethics of the profession of those at the FDA doesn’t allow certain things. There is no way any group of reviewers would cross any boundaries. The regulation is really very tight,” she says. Besides, “every year, you get checked” through audits and government investigations.

Kaitin says it’s all part of the important public debate about where society is willing to set its “risk bar” on the scale of a drug’s benefits and risks. “Congress and the FDA react on each other” to create balanced policy, he says.

PDUFA, which came in the wake of AIDS activists protesting the turtle-paced drug approvals in the late 1980s when there was little to offer dying patients, tipped the scale in favor of quick approvals and increased risk. “With the passage of PDUFA, the view of the FDA changed dramatically,” Kaitin says. Prior to 1992, he says CDER looked at an application for reasons not to approve it. Today, when an application is submitted, reviewers consider that it should be approved, unless something signals that it shouldn’t. “The reason why is this level of trust…because the agency and the industry have agreed to work more closely.”

But he says he can see the scale tipping the other way now. “There’s no one pressing for a less rigorous approval of drugs.” The AIDS activists who protested FDA policies in an effort to get some drugs—any drugs—on the market two decades ago aren’t standing in front of the agency’s suburban Maryland buildings anymore. Approval times have recently inched back up, and while he says current FDA structure doesn’t make a pre-PDUFA process possible, he applauds McClellan—with his M.D. and Ph.D. in economics—for having achieved balance between the two sides. “What Mark McClellan did…he managed to reframe the debate…. He put a public face on it.”

But McClellan couldn’t remove the agency from political infighting. Waxman says this is a problem, although he clearly is taking sides in the battle. “The Bush administration’s willingness to place ideology ahead of science” caused the agency to lose credibility, he says. “It hurts all the science-based agencies when decisions are made using politics.” He points to the FDA’s recent postponement of a decision regarding Plan B, despite its advisory committee approval, calling it a political move. A final decision is expected this month.

With politics weighing so heavily on the FDA and a new commissioner in the distant horizon, no one can say for sure where the scale will settle in this ongoing debate. In the meantime, the work piles up at the FDA. How it continues to go from a binder on an 18-wheeler to a bottle behind the counter will surely be a function of both politics and professionalism.
Jennifer Zeigler is a senior writer with The New Physician. Direct comments about this article to