Can FDA Reform Drug Reviews?

Date: March 16, 1998 New Drug Development A regional lab visit highlights challenges of agency-wide change At the Food and Drug Administration's Central Region office in Philadelphia, cabinetsful of documentation offer only a vague notion of the paperwork involved in getting a new drug through FDA's review and approval process. Nicholas Falcone, the district office's coordinator of new drug applications (NDAs) and abbreviated new drug applications (ANDAs)-the latter for generic compounds-hefts two tomes bound in heavy stock. These, he says, represent just over one volume of a typical application that runs to 150 volumes. Multiple copies are made of each application, he points out. "If you're ever at FDA headquarters in Rockville [Md.] to see a delivery truck come in, you'll know what I mean." The FDA Modernization Act (FDAMA) of 1997, signed into law last November, provides funds for electronic applications, but the agency's district outposts will keep receiving paperwork for the foreseeable future, Falcone acknowledges. LAB WORK: New drug applications coordinator Nick Falcone, left, and chemist Nasir Ali check an HPLC analysis in Philadelphia. These masses of paper symbolize a central FDA challenge nowadays: to streamline its approval process for new drugs and biologics. It's too early after enactment of FDAMA to determine whether or not implementation of the law's many provisions will help to speed approval time and reduce the cost of drug research and development. Within industry, both support for the new legislation and criticism of it remain strong, indicating the hard-fought nature of the negotiations that preceded its enactment. Well away from the seat of power, at the district office in Philadelphia, can any changes be detected yet that might bode well for speedier drug approvals? On the plus side, the Philadelphia office-which employs 126 people, of whom 29 are lab workers-underwent a $1.3 million upgrade last year. But on the other hand, FDA is reacting to budgetary pressures of recent years by scaling back its laboratories nationwide, a plan announced by Gary Dykstra, deputy associate commissioner for regulatory affairs, at the joint Pharm Tech and BioPharm Conferences held in Philadelphia September 15-17. He said the then-17 FDA labs would shrink to nine over coming years, only two of which would concentrate entirely on drug work-one in San Juan, Puerto Rico, and the other in Philadelphia (Notebook, The Scientist, 11[21]:31, Oct. 27, 1997). Since then, three FDA labs have shut down. The Philadelphia office's science branch director, W. Charles Becoat, says, "As the labs close, we'll get more people and we'll expand. The goal is for us to do the bulk of the NDA and ANDA work here." Like the paperwork, the chemistry done at the Philadelphia office symbolizes another key question concerning whether the drug approval process can be streamlined: What roles should FDA personnel be filling? The latest point of contention in this regard is over agency plans to belt-tighten by shedding almost one-third of its nearly 280 researcher/reviewers in the Center for Biological Evaluation and Research (CBER) within the next two years (B. Agnew, Science, 279:976-7, 1998). CBER oversees a subset of therapeutics that includes vaccines, extracts of living cells, certain extracts of natural sustances, blood, and blood products. Its sister organization, the Center for Drug Evaluation and Research (CDER), is responsible for both prescription and over-the-counter drugs for humans (the other centers regulate food and nutrition, medical and radiologic devices, and veterinary medicine). But CBER is the only center that combines the roles of researcher and reviewer. GROWTH EXPECTED: As other FDA labs close across the country, the Philadelphia lab's work will increase, says manager Charles Becolt. Within industry, a cutback in CBER researchers has both supporters and detractors. "The quality of people in FDA is much higher than some people say," declares George B. Rathmann, chairman and CEO of Bothell, Wash.-based ICOS Corp., a publicly held company that researches and develops new pharmaceuticals for chronic inflammatory diseases. He believes skilled CBER researcher/reviewers are helpful, because the chemical structure of a biologic often has a more clear developmental "logic" than that of a pharmaceutical. This means if the potential end use of a biologic changes during development from, say, a treatment for inflammatory bowel disease to a treatment for asthma, the same group of researcher/reviewers at CBER will still be working with the biologic's sponsor. At CDER, new reviewers would take over the project, thus complicating the industry partner's collaboration with the agency. The Philadelphia lab does not analyze biologics, and its work on new drugs is at the very end of the approval process, Becoat is careful to point out. By the time his team gets its slice of the paperwork in an application, years of drug discovery and development, of animal and clinical trials, have been completed by the drug's sponsor. About 70 percent of the analysis done in Philadelphia is high-performance liquid chromatography to determine that the composition and amounts of the compound meet specifications contained in the application documentation. However, some of the chemistry done in Philadelphia potentially has implications for standards-setting. For example, chemist Syed Nasir Ali uses a dissolution apparatus to simulate the ability of a human stomach to dissolve a given drug. For generic products, this process also helps to differentiate one company's drug from the same drug made by another company. The studies give Ali statistics on dissolution that sometimes prompt him to propose changes to a drug's "public standards," contained in the official United States Pharmacopoeia. "There's no question that researchers doing cutting-edge stuff [are] useful at FDA," says Henry I. Miller, a senior research fellow at the Hoover Institution of Stanford University, and director from 1989 to 1993 of FDA's office of biotechnology. "The question is whether it's cost-effective or not." At least once every two years, all of the approximately 1,500 United States establishments that manufacture, test, pack, and label drug products are inspected by Food and Drug Administration personnel. At FDA's district office in Philadelphia, preapproval program manager Deborah L. Pagano and 59 inspectors ensure that drug manufacturing facilities in their region can produce to quality standards a new drug whose application is being considered for approval by reviewers at the agency's headquarters. Preapproval inspectors examine equipment, technology transfer, lab procedures, analytical methods, and documentation. A primary objective is to determine if raw data submitted in the application are authentic and accurate. For example, data in bound notebooks with pages ripped out are immediately suspect. Or, says Pagano, "You don't want to see people shaking a solution as they go down the hall so that it's more mixed." Pagano is pleased with new software piloted in the Philadelphia office early in 1997 and online since last September in all 21 district offices. "If a reviewer types in a comment on an application, I now have access to it," she says. Previously, she spent a lot more time on the telephone, trying to discover who the reviewers were at headquarters for any given drug, and what they thought. FDA's Center for Drug Evaluation and Re-search has about 900 reviewers-half its work force-including chemists, pharmacologists, physicians, pharmacokineticists, statisticians, and microbiologists. The reviewers file the requirements for each firm to comply with regulations. It's an unwritten rule, says Pagano, that the reviewer can decide on a new drug application without any input at all from the district office. But she adds, that doesn't happen. "I would say 99 percent of the time, we coincide with what the reviewer is going to recommend."-S.B. Industry has several reasons to be interested in that question. One reason is that the average cost in the U.S. of bringing a new drug through the approval process to market has been estimated at $500 million by the Boston Consulting Group. Another reason is that FDA's only source of operating income not derived from federal government coffers comes from "user fees" paid by companies for review of new drug and biologics applications. About $132 million in user fees was paid in fiscal year 1997, the last year before expiration of the Prescription Drug User Fee Act (PDUFA) of 1992. These funds enabled FDA to add 600 new drug reviewers and support staff, thus helping to shorten the time required for approval before companies may market their new products. In 1997, when FDA was named one of 10 annual winners of the Innovations in American Govern-ment Awards Program-sponsored by the Ford Foundation and Harvard University's John F. Kennedy School of Government, in partnership with the Council in Excellen ce in Government-a press re-lease attributed the agency's award to its "streamlined drug review process." The release declared, "U.S. drug approval times have decreased dramatically and are now among the fastest in the world." Rathmann disagrees. He says FDA's drug approval process was "much faster in the 1980s than it is now," and he decries the agency's "intractable delays, due to precautions that many times are totally unnecessary or worthless." Under FDAMA, the agency's performance goals include decreasing the time for "standard review" from its current 12 months to 10 months. However, Rathmann charges that this is merely the time required by FDA to review NDA documentation. "They have speeded up the last stage and slowed down everything " he says. "They have resisted the much more difficult problem of accelerating the approval cycle that begins with the IND [Investigational New Drug application]." The IND application must be submitted to FDA by the drug's sponsor before the start of clinical trials. This paperwork includes the plan for human studies, information on the drug's structural formula, animal test results, and manufacturing information. Rathmann says the years required to complete the process are growing rather than decreasing. Although he lauds codification in FDAMA that one well-controlled clinical trial may be deemed sufficient to establish a new drug's effectiveness, he also attributes longer drug development times largely to FDA's increased scrutiny of clinical trials in recent years. He points out that his company-which has collaborative agreements with Glaxo Wellcome Inc. of Research Triangle Park, N.C, United Kingdom-based Cambridge Antibody Technology, and Abbott Laboratories of Abbott Park, Ill.-spends $50 million annually on drug research and development and is in its eighth year of existence but has not yet gotten a drug on the market. "By being a watchdog that's overly active, [FDA is] doing a disservice to patients." The Washington, D.C.-based Competitive Enterprise Institute (CEI), a nonprofit public-policy group that describes its mission as "advancing the principles of free enterprise and limited government," compiled a monograph in 1997 (J.C. DeFalco, CEI Monograph, February 1997) on certain drugs approved in 1996 by FDA. The report reflects unfavorably on FDA's approval times for these drugs-to treat breast cancer, pancreatic cancer, epilepsy, interstitial cystitis, and brain cancer-compared with approval times in other countries. Miller warns that head-to-head comparisons between approval times for the same drug in different countries are difficult to do accurately because of the many variables involved. Nevertheless, he says, about three-quarters of drugs that receive FDA approval have already been approved somewhere in the world. "Given that, FDA's numbers should be better than here, and not worse." He says in the U.S., the average time required to take a drug from research lab to patient's bedside has more than doubled since 1964, from 6.5 to 14.8 years (H.I. Miller, The Scientist, 11[24]:9, Dec. 8, 1997). Figures from Tufts University's Tufts Center for the Study of Drug Development show an upward trend in total development time and in clinical trial time for new drugs since the 1960s, although drug application approval time is at an historic low (J.A. DiMasi, Chart 5-8 in Economic Report of the President, February 1998, page 189). Among FDAMA's strong supporters within industry is the Washington, D.C.-based Biotechnology Industry Organization (BIO), which represents more than 750 companies, academic institutions, and state biotechnology centers in the U.S. and overseas. BIO was an energetic participant in negotiations over the legislation, which began in earnest in mid-1994. It was a "politically volatile, polarized environment," recalls Carl Feldbaum, BIO president. "It was a battle between what I called the ideologues in Congress and the theologues in FDA, and that proved to be too much of a chasm to cross." However, he adds, "At the end of 1995, biotechnology CEOs and regulatory experts began a s eries of conversations with [then-FDA commissioner] David Kessler, which proved to be eye-opening and door-opening on both sides." Feldbaum believes that FDAMA heralds "a whole new atmosphere of cooperation, with a clear understanding of the roles of the different parties." But Feldbaum, Rathmann, Miller, and others agree that Michael Friedman's continuing role as FDA's acting commissioner following Kessler's departure early last year has left the agency without a critical sense of permanent leadership. As Feldbaum puts it, implementation of the many measures in FDAMA that are designed to streamline the drug approval process is "partly contingent on the appointment of a new FDA commissioner who shares these goals wholeheartedly-and we would like to see that person named." New Drug Development

Can FDA Reform Drug Reviews?

Interested in reading more?

Become a Member of