Funding the fast track

ILLUSTRATION: GEOFF SMITH

New drug approvals in the United States today are handled by a chronically short-staffed, overworked federal agency that somehow manages to process dozens of applications annually (some more than 100,000 pages long) while trying to accommodate profound changes in health care delivery and the new challenges of a Web-based economy.

The review process
The U.S. Food and Drug Administration’s Center for Drug Evaluation and Research (CDER) was created in 1987, when the existing Center for Drugs and Biologics was split into two separate divisions: CDER and the Center for Biologics Evaluation and Research, the latter being largely responsible for vaccines. The change was considered necessary at the time to handle the increasing volume of New Drug Applications (NDAs) and to provide high-level attention for the growing problem of AIDS.

Today, CDER’s process for handling approvals depends on whether the drugs are prescription, generic, or over-the-counter (OTC). For each category, the process is substantially different, according to Sandra Kweder, acting director of CDER’s Office of Review Management. OTC reviews —handled by a single dedicated division in the agency—can involve more legal compliance than scientific review. “We’ve assembled a list of guidelines, known as monographs, that describe specific OTC formulations for treating specific conditions,” she explains. “As long as the drug company complies, the drugs are approved.” Entirely new OTC drugs, she adds, are given a full review similar in scope to that reserved for prescription products.

Generic drug reviews, also the purview of a dedicated division, are relatively straightforward because drug companies merely have to show that the generic product is bioequivalent to the parent or “innovator” compound. Other stipulations require that the generic drug can be manufactured in a “reproducible manner under controlled conditions,” and that the information on the generic and innovator labels is essentially identical.

The most intensive evaluations by far are those for new prescription drugs. Prescription NDAs are handled by internal-review divisions organized around 15 specialty areas, such as oncological or pulmonary drug products. The review teams themselves comprise 10–15 agency experts in chemistry, pharmacology, medicine, pharmacokinetics, statistics, and microbiology. According to Kweder, the agency will sometimes consult outside experts, particularly about decisions that involve “close calls”, through its Office of Advisors and Consultants. “When we need to, we submit certain details pertaining to panels of external experts who can advise us on the matter at hand,” she says. “But they don’t do the actual review—we bring the review to them.”

For an NDA to be counted among the 30–40% that make it on the first try, companies must follow detailed guidance contained in 21 CFR Part 312 (www.fda.gov/cder/regulatory/applications/ind_page_1.htm) and provide a complete package that holds up to an intense level of agency scrutiny. The required documentation tells a drug’s story: history and data pertaining to clinical trials, components and composition, results of animal toxicologic and pharmacokinetic studies, and plans for manufacturing, processing, and packaging. If the NDA fulfills a manufacturer’s claims without excessive side effects—those of greater magnitude than described in the proposed label—the review can be completed in 10–12 months. High-priority drugs, including all AIDS medications and others that offer a significant medical advance over existing therapies, can be processed in 6 months or less.

In the final analysis, CDER’s approval decision boils down to two criteria: whether the results of well-controlled studies provide sufficient evidence of effectiveness, and whether studies show that the drug is safe for use under the conditions described in the proposed label. If these conditions are met, CDER notifies the applicant that the drug is approved for marketing. If not, CDER returns the NDA with an explanation of why the product has failed the review.

Too slow?
The Prescription Drug User Fee Act (PDUFA), passed in 1992, allowed CDER to begin collecting fees from companies that submit NDAs for agency review. The subsequent influx of corporate money allowed CDER to hire more than 500 additional reviewers—a swelling of the ranks that doubled the number of annual new drug approvals while cutting review times nearly in half. The PDUFA is reauthorized every five years, and CDER is awaiting a third round of legislative authority under the act, which is scheduled for 2002.

Recently, however, CDER has been criticized by those who think it is dragging its feet on new drug approvals. And in fact, the number of new drugs released to the market annually is falling steadily, from a high of 161 in 1996 to only 38 in 2000, according to IMS Health. Industry analysts suggest that this trend has had many impacts, including a drop in the amount of money available for basic research, and an enticement for pharmaceutical companies to spend excessive sums on marketing for their currently approved products.

The cause of the slowdown remains somewhat unclear. Alan Goldhammer, vice president for regulatory affairs at the Pharmaceutical Research and Manufacturers of America, suggests that with the additional reviewers hired after PDUFA, CDER was able to plow through a backlog of NDAs until approvals peaked in 1996. The subsequent drop in drug approvals, he says, could simply reflect the elimination of this pervasive waiting list. Wayne Pines, the former associate commissioner of the FDA and currently president of health care at APCO Worldwide (Washington, DC), links the trend at least in part to agency caution after some controversial drug withdrawals made in the late 1990s. “A number of drugs—for example, fenfluramine, which was found to produce adverse cardiac reactions; Resulin; and Posicor—were pulled from the market because of unexpected adverse reactions,” he says. “So CDER became more conservative, not necessarily slower, but it began to demand more data in the NDA.”

Kweder disputes the notion that more products were withdrawn in the late 1990s than at other times. She says that the rate of withdrawals has been “flat for decades,” averaging roughly 2–3% a year. The number of patients evaluated during clinical trials limits the agency review of side effects during the NDA process. The conclusive test of a drug’s safety actually takes place in the marketplace, where adverse reactions can be measured in populations numbering in the tens of thousands or higher. (CDER’s responsibility for monitoring these side effects continues throughout the drug’s marketed lifespan.) Some stakeholders suggest that media coverage accompanying the withdrawal of fenfluramine, which was used in several high-profile weight-loss combinations, may explain an excessive level of agency caution. But again, Kweder is reluctant to accept this possibility.

“People ask us about that all the time,” she says. “It’s hard to quantify whether or not there’s any more caution at the agency. Sometimes we don’t become aware of a particular side effect until the drug has already been released to the market. We have a responsibility to take a hard look at these emerging side effects when we make review decisions.”

According to Kweder, the slowdown in new approvals may have more to do with fewer applications from the drug companies, in addition to strains on review teams that leave CDER struggling to keep its head above water. “Our reviewers are stretched very thin,” she says. “[Legislative changes] during the 1990s tripled the workload related to deliverables while providing very little in the way of additional resources. Right now, we’re trying to fix that by negotiating for additional funding under the third round of PDUFA authorization.”

Ultimately, concerns over delayed approvals are valid because companies need to have a dependable review process, says Pines. “To make research investments and establish appropriate priorities, it serves everybody’s interests to have a predictable and consistent FDA,” he says.

Toward the future
Another challenge for CDER as it moves ahead is adjusting to a marketplace increasingly dominated by managed care. When it comes to new drug approvals, the agency’s mandate has always been to evaluate a drug’s safety and efficacy, not its cost. But in a managed health care environment, providers are increasingly forced to consider economic factors when making treatment decisions. Perhaps the best source of comparative information that might enable clinicians to choose lower-priced drugs from their alternatives is the content of CDER’s reviews. Drug labels produced by the manufacturers are less useful in this respect. Says Pines, “The fundamental question for managed care is whether the new drug is worth the money. And for that kind of judgment, you really have to go back to the FDA.”

Does CDER provide the information to those who seek it? Yes. But according to Kweder, the traditional route to the data is through a formal, cumbersome request made under the Freedom of Information Act. “Our reviews are written by doctors, and they show precisely how a drug fits into the whole treatment,” she says. “We have to find better ways of making this information available to the public. We are cognizant of the need to communicate with the outside world.”

As for the entire medical–industrial complex, the Internet will probably play a critical role as developments in the drug approval process unfold. CDER is already making substantial headway toward a completely electronic data submission and review environment, which is scheduled to emerge in 2002. Today, the paperwork associated with the typical NDA is so extensive that it is transported in trucks and moved with forklifts. With electronic submissions, the whole dataset will be carried on one or two CD-ROMs. Progress, in this case at least, will be measured in smaller packages.