What Ails the FDA?
http://www.100md.com
《新英格兰医药杂志》
Testifying before a spellbound audience in a Senate hearing room this past November, scientist and whistle-blower David J. Graham charged that his employer, the Food and Drug Administration (FDA), is incapable of protecting the public from dangerous prescription drugs. Graham, an epidemiologist who monitors drug safety, conducted a study that confirmed an increase in the risk of myocardial infarction among users of rofecoxib (Vioxx), Merck's blockbuster antiinflammatory medicine. When he informed his supervisors about his findings last summer, he said, they pressured him to change his conclusions because they were inconsistent with the FDA's position on the drug's safety. A few weeks later, Merck withdrew Vioxx from the market in the face of studies suggesting that it had contributed to tens of thousands of heart attacks.
"Vioxx is a terrible tragedy and a profound regulatory failure," Graham told the Senate Finance Committee. "I would argue that the FDA, as currently configured, is incapable of protecting America against another Vioxx. We are virtually defenseless."
The Vioxx debacle and the FDA's slowness in warning physicians and patients about studies linking the use of antidepressants among adolescents to an increased risk of suicidal behavior have raised questions about the agency's ability to fulfill one of its fundamental missions — to ensure that the benefits of prescription drugs outweigh their risks. The FDA's recent performance on drug safety has drawn fire from members of Congress, consumer advocates, and experts within the medical and public policy communities. In particular, there is broad agreement that we need better ways to monitor the safety of drugs once they are on the market, a problem that is to be examined by an Institute of Medicine committee.
Yet the need for better surveillance of approved drugs is only one of the FDA's ills. Another is the lack of strong and independent leadership, which has contributed to an atmosphere that stifles debate and discourages some employees from expressing scientific concerns about drugs. This leadership vacuum, combined with both the pressure to approve new products quickly and the antiregulatory philosophy of the current administration, has created an FDA that seems timid and toothless.
For most of the time that President George W. Bush has been in office, the FDA has lacked a permanent leader. Bush's first appointee as FDA commissioner, Mark McClellan, was confirmed in November 2002 and left in March 2004 to become the administrator of the Centers for Medicare and Medicaid Services. For more than a year before McClellan's confirmation, the FDA's highest-ranking appointed administrator was chief counsel Daniel Troy, who had represented tobacco and drug companies in suits fighting the FDA's proposed regulation of tobacco and opposing its efforts to restrict the promotion of drugs for unapproved uses.
The current acting FDA commissioner, Lester Crawford, served as acting commissioner for a total of almost three years, during periods before and after McClellan's tenure, before being nominated last month to become the agency's permanent head. The deputy commissioner, Janet Woodcock, also serves in an acting capacity, as does Steven Galson, director of the Center for Drug Evaluation and Research (CDER), which is responsible for approving new drugs and ensuring the safety and effectiveness of all drugs. And so does Graham's boss, Paul Seligman, acting director of the Office of Drug Safety, whose staff of pharmacologists and epidemiologists monitor adverse-event reports on drugs, watching for "signals" that suggest safety problems.
The Office of Drug Safety has no role in determining whether drugs are safe before they are approved. This is the task of review teams within CDER's Office of New Drugs. Each team is led by a project manager and includes a medical officer (a clinically trained physician), a chemist, a pharmacologist, and other members. The review team discusses with company scientists the design of clinical studies to be submitted in applying for approval of a new drug — including their size, duration, and patient population. The team decides the amount of data that are adequate to determine the acceptability of a drug's risk–benefit ratio. Once a drug has been approved, the FDA has no legal authority to require additional safety studies, although it can ask a manufacturer to include safety measures in any further trials that are done to confirm efficacy or support a new indication.
In 1992, Congress passed the Prescription Drug User Fee Act (PDUFA) in response to complaints from the pharmaceutical industry and the medical community about the FDA's slow pace in reviewing applications for new drugs and biologic products. Under PDUFA, companies began to pay fees to the agency, which were to be used to hire more reviewers and make other changes in order to speed up the approval process. These fees were contingent on the FDA's adherence to strict review timetables, and until recently, none of this money could be spent for other activities. (Since reauthorization of PDUFA in 2002, use of some PDUFA funds for the FDA's drug-safety program has been permitted.) Between 1993 and 2002, user fees allowed the FDA to increase by 77 percent the number of personnel assigned to review applications, according to a report issued by the General Accounting Office (GAO) in September 2002. User fees from pharmaceutical companies now account for more than half the money the FDA spends on the review process.
And the agency has, in fact, dramatically shortened its time to approval for new drug applications (see graph). For drugs given a "priority review" (because they were judged to offer a therapeutic advantage over existing medications), the median time to approval decreased from 14.9 months in 1993 to 6.7 months in 2003. For drugs given a "standard review," the median time fell from 27.2 months to 23.1 months during the same period. "We think that our pre-market process is pretty good," Deputy Commissioner Woodcock noted. "We think that PDUFA has done a great deal for bringing better science and more reviewers to the pre-market review."
Median Total Time to FDA Approval for New Drug Applications, 1986–2003.
The Prescription Drug User Fee Act was passed in 1992, creating an accelerated ("priority") review process that is paid for by user fees from pharmaceutical companies. The number at the bottom of each bar is the number of drugs reviewed. Data are from the Center for Drug Evaluation and Research.
But other observers believe that the changes brought by PDUFA have impaired reviewers' ability to assess drug safety impartially by fostering a frenetic atmosphere in which the pharmaceutical industry is viewed as the customer and scientific debate is discouraged. The GAO report concluded that PDUFA had forced the FDA to shift funds away from other activities, including post-marketing safety surveillance, and had contributed to increased workload, high turnover rates, and reduced training time for scientists and medical officers on review teams. In a survey of CDER scientists conducted in 2002 by the Office of Inspector General of the Department of Health and Human Services, 18 percent of the respondents said that they had "been pressured to approve or recommend approval" of a drug despite having reservations about its safety, efficacy, or quality. For drugs assigned to priority review, 58 percent of the respondents said that reviewers were not given enough time "to conduct an in-depth, science-based review."1
"We were on a clock," Elizabeth Barbehenn said in an interview. A pharmacologist who left the FDA in 1998 after 13 years on a CDER drug-review team, Barbehenn said, "We had just so much time to get a review done. I found it extraordinarily frustrating."
After PDUFA took effect, Barbehenn said, she was allowed to communicate only indirectly, through a supervisor, with scientists at the companies whose drugs she was reviewing. "You really have to know the science well to argue with industry," she said. "Your supervisor, with so many more drugs to deal with, couldn't really keep on top of anything in the same way. . . . It's very much driven by what industry wants." Another scientist, who has worked at the FDA for about 15 years, said in an interview that PDUFA produced a "sea change" in the priorities set by agency managers. "When I joined, there was an absolute emphasis on safety," he said. "It is very, very clear that the emphasis now is getting drugs approved. To justify not getting them approved is considerably more difficult."
Alastair J.J. Wood, a professor of pharmacology and medicine at Vanderbilt University School of Medicine who has served on FDA advisory committees since the early 1990s, told me that in recent years he has encountered increased reticence on the part of the agency's medical officers and scientists. "There's certainly more of a culture of fear," he said. "They're much more cautious about being perceived as critical of the hierarchy or the system or the leadership."
In addition to reviewing drugs more rapidly, in recent years FDA review teams have in many cases based their decisions on insufficient evidence, according to Jerry Avorn, a professor of medicine at Harvard Medical School and author of Powerful Medicines (2004), on prescription drugs. Avorn said in an interview that companies are frequently allowed to submit pivotal studies that lasted only a few months, even for drugs that will be taken for a long time, and to recruit subjects who are not representative of the population likely to take a drug after it is marketed. The agency has also been increasingly willing to grant "accelerated approval" on the basis of a drug's effect on a surrogate marker, such as tumor shrinkage for a cancer drug or reduction of low-density lipoprotein cholesterol for a statin. Federal rules require companies receiving accelerated approval to conduct post-marketing studies that use more definitive outcome measures, but more often than not, companies fail to do so — and the FDA has never responded to such lapses by withdrawing a drug. Of more than 1300 post-marketing studies to which drug companies have committed themselves, 65 percent have not been started.2
David A. Kessler, who was FDA commissioner when this program began, said that PDUFA was meant to provide the FDA with the necessary resources to speed up reviews, not to lower the bar for approval. He added that during his tenure he noticed no lowering of standards. "PDUFA has nothing to do with debate and discussion" within the FDA, he said. "That's leadership. That's the culture of the agency."
Have faster reviews by the FDA contributed to an increase in the number of drugs withdrawn from the market? The answer isn't clear. In the 2002 GAO report, investigators noted an increase in the rate of safety-related drug withdrawals (as expressed by the ratio of the number of drugs withdrawn for safety reasons each year to the number of drugs approved). From 1997 through 2000, the withdrawal rate was 5.34 percent, as compared with 1.96 percent between 1989 and 1992 (pre-PDUFA) and 1.56 percent between 1993 and 1996 (immediately after the enactment of PDUFA). But a recent FDA analysis, which examined drug withdrawals according to year of approval, showed little change after PDUFA was enacted. From 1971 through 1993, the rate of safety-based drug withdrawals was 2.7 percent, as compared with 2.3 percent for drugs approved between January 1, 1994, and April 30, 2004.3
Safety-related withdrawals occur sporadically and for varying reasons — for example, there were five such withdrawals between September 1997 and September 1998, but none in 2003. Recently withdrawn drugs have included one, fenfluramine, that was approved in 1973 after a 75-month review and two others, troglitazone and cerivastatin, that were approved in 1997 after reviews lasting 6 months and 12 months, respectively. Rofecoxib is among the most swiftly approved drugs to be withdrawn for safety reasons. Considered potentially safer for the gastrointestinal tract than older nonsteroidal antiinflammatory drugs, it received a priority six-month review.
Among the most worrisome signs that things are amiss within the agency are reports that FDA scientists have been discouraged by supervisors from raising questions about drug safety and sometimes have been prevented from sharing their concern with FDA advisory committees. In February 2004, medical reviewer Andrew Mosholder was prevented from presenting to an advisory panel his analysis linking the use of antidepressant drugs in teenagers with an increased risk of suicidal behavior; instead, his superiors ordered a second analysis by another FDA scientist, which reached the same conclusion six months later.4 Barbehenn, the former FDA pharmacologist, said that medical officers on her CDER review team were prevented by supervisors from presenting an analysis on alendronate (Fosamax) at an advisory committee meeting. Graham said in an interview that he and other scientists from the Office of Drug Safety sometimes attend advisory-committee meetings but are rarely invited to present data or analyses.
This past summer, faced with supervisors' demands that he amend the conclusions of his Vioxx study, Graham did so, he told the Senate Finance Committee, in order to obtain permission to present the findings at an international conference. "I changed them to a fair degree, and it caused me a great deal of mental anguish," he said. "I did it because I thought if I didn't, there was no way that the data would see the light of day."
Last month, Health and Human Services Secretary Mike Leavitt announced plans to establish a new Drug Safety Oversight Board within the FDA that will draw on outside expertise to review safety issues that arise with regard to approved drugs. Graham's supervisors also permitted him to present findings from an unpublished study to the FDA advisory committee reviewing the safety of cyclooxygenase-2 (COX-2) inhibitors — a decision for which Graham publicly thanked Crawford. These are encouraging signs of change. But some experts have suggested that to achieve impartial scientific assessments of risks, the drug-safety program should be moved out of the FDA into an independent agency. "You can't run a safety system where you appear to be suppressing conflicting opinions, even if these conflicting opinions turn out to be wrong," said Vanderbilt's Wood.
Source Information
Dr. Okie is a contributing editor of the Journal.
References
Office of Inspector General. FDA's review process for new drug applications: a management review. March 2003. (OEI-01-00590.) (Accessed February 24, 2005, at http://www.oig.hhs.gov/oei/reports/oei-01-01-00590.pdf.)
Food and Drug Administration. Report on the performance of drug and biologics firms in conducting postmarketing commitment studies: availability. Fed Regist 2004;69:12162-12164.
Food and Drug Administration. CDER report to the nation: 2003. (Accessed February 24, 2005, at http://www.fda.gov/cder/reports/rtn/2003/rtn2003-3.htm.)
Vendantam S. FDA study confirms antidepressant risks: drugs linked to more suicides among children, unpublished analysis says. Washington Post. August 10, 2004:A6.(Susan Okie, M.D.)
"Vioxx is a terrible tragedy and a profound regulatory failure," Graham told the Senate Finance Committee. "I would argue that the FDA, as currently configured, is incapable of protecting America against another Vioxx. We are virtually defenseless."
The Vioxx debacle and the FDA's slowness in warning physicians and patients about studies linking the use of antidepressants among adolescents to an increased risk of suicidal behavior have raised questions about the agency's ability to fulfill one of its fundamental missions — to ensure that the benefits of prescription drugs outweigh their risks. The FDA's recent performance on drug safety has drawn fire from members of Congress, consumer advocates, and experts within the medical and public policy communities. In particular, there is broad agreement that we need better ways to monitor the safety of drugs once they are on the market, a problem that is to be examined by an Institute of Medicine committee.
Yet the need for better surveillance of approved drugs is only one of the FDA's ills. Another is the lack of strong and independent leadership, which has contributed to an atmosphere that stifles debate and discourages some employees from expressing scientific concerns about drugs. This leadership vacuum, combined with both the pressure to approve new products quickly and the antiregulatory philosophy of the current administration, has created an FDA that seems timid and toothless.
For most of the time that President George W. Bush has been in office, the FDA has lacked a permanent leader. Bush's first appointee as FDA commissioner, Mark McClellan, was confirmed in November 2002 and left in March 2004 to become the administrator of the Centers for Medicare and Medicaid Services. For more than a year before McClellan's confirmation, the FDA's highest-ranking appointed administrator was chief counsel Daniel Troy, who had represented tobacco and drug companies in suits fighting the FDA's proposed regulation of tobacco and opposing its efforts to restrict the promotion of drugs for unapproved uses.
The current acting FDA commissioner, Lester Crawford, served as acting commissioner for a total of almost three years, during periods before and after McClellan's tenure, before being nominated last month to become the agency's permanent head. The deputy commissioner, Janet Woodcock, also serves in an acting capacity, as does Steven Galson, director of the Center for Drug Evaluation and Research (CDER), which is responsible for approving new drugs and ensuring the safety and effectiveness of all drugs. And so does Graham's boss, Paul Seligman, acting director of the Office of Drug Safety, whose staff of pharmacologists and epidemiologists monitor adverse-event reports on drugs, watching for "signals" that suggest safety problems.
The Office of Drug Safety has no role in determining whether drugs are safe before they are approved. This is the task of review teams within CDER's Office of New Drugs. Each team is led by a project manager and includes a medical officer (a clinically trained physician), a chemist, a pharmacologist, and other members. The review team discusses with company scientists the design of clinical studies to be submitted in applying for approval of a new drug — including their size, duration, and patient population. The team decides the amount of data that are adequate to determine the acceptability of a drug's risk–benefit ratio. Once a drug has been approved, the FDA has no legal authority to require additional safety studies, although it can ask a manufacturer to include safety measures in any further trials that are done to confirm efficacy or support a new indication.
In 1992, Congress passed the Prescription Drug User Fee Act (PDUFA) in response to complaints from the pharmaceutical industry and the medical community about the FDA's slow pace in reviewing applications for new drugs and biologic products. Under PDUFA, companies began to pay fees to the agency, which were to be used to hire more reviewers and make other changes in order to speed up the approval process. These fees were contingent on the FDA's adherence to strict review timetables, and until recently, none of this money could be spent for other activities. (Since reauthorization of PDUFA in 2002, use of some PDUFA funds for the FDA's drug-safety program has been permitted.) Between 1993 and 2002, user fees allowed the FDA to increase by 77 percent the number of personnel assigned to review applications, according to a report issued by the General Accounting Office (GAO) in September 2002. User fees from pharmaceutical companies now account for more than half the money the FDA spends on the review process.
And the agency has, in fact, dramatically shortened its time to approval for new drug applications (see graph). For drugs given a "priority review" (because they were judged to offer a therapeutic advantage over existing medications), the median time to approval decreased from 14.9 months in 1993 to 6.7 months in 2003. For drugs given a "standard review," the median time fell from 27.2 months to 23.1 months during the same period. "We think that our pre-market process is pretty good," Deputy Commissioner Woodcock noted. "We think that PDUFA has done a great deal for bringing better science and more reviewers to the pre-market review."
Median Total Time to FDA Approval for New Drug Applications, 1986–2003.
The Prescription Drug User Fee Act was passed in 1992, creating an accelerated ("priority") review process that is paid for by user fees from pharmaceutical companies. The number at the bottom of each bar is the number of drugs reviewed. Data are from the Center for Drug Evaluation and Research.
But other observers believe that the changes brought by PDUFA have impaired reviewers' ability to assess drug safety impartially by fostering a frenetic atmosphere in which the pharmaceutical industry is viewed as the customer and scientific debate is discouraged. The GAO report concluded that PDUFA had forced the FDA to shift funds away from other activities, including post-marketing safety surveillance, and had contributed to increased workload, high turnover rates, and reduced training time for scientists and medical officers on review teams. In a survey of CDER scientists conducted in 2002 by the Office of Inspector General of the Department of Health and Human Services, 18 percent of the respondents said that they had "been pressured to approve or recommend approval" of a drug despite having reservations about its safety, efficacy, or quality. For drugs assigned to priority review, 58 percent of the respondents said that reviewers were not given enough time "to conduct an in-depth, science-based review."1
"We were on a clock," Elizabeth Barbehenn said in an interview. A pharmacologist who left the FDA in 1998 after 13 years on a CDER drug-review team, Barbehenn said, "We had just so much time to get a review done. I found it extraordinarily frustrating."
After PDUFA took effect, Barbehenn said, she was allowed to communicate only indirectly, through a supervisor, with scientists at the companies whose drugs she was reviewing. "You really have to know the science well to argue with industry," she said. "Your supervisor, with so many more drugs to deal with, couldn't really keep on top of anything in the same way. . . . It's very much driven by what industry wants." Another scientist, who has worked at the FDA for about 15 years, said in an interview that PDUFA produced a "sea change" in the priorities set by agency managers. "When I joined, there was an absolute emphasis on safety," he said. "It is very, very clear that the emphasis now is getting drugs approved. To justify not getting them approved is considerably more difficult."
Alastair J.J. Wood, a professor of pharmacology and medicine at Vanderbilt University School of Medicine who has served on FDA advisory committees since the early 1990s, told me that in recent years he has encountered increased reticence on the part of the agency's medical officers and scientists. "There's certainly more of a culture of fear," he said. "They're much more cautious about being perceived as critical of the hierarchy or the system or the leadership."
In addition to reviewing drugs more rapidly, in recent years FDA review teams have in many cases based their decisions on insufficient evidence, according to Jerry Avorn, a professor of medicine at Harvard Medical School and author of Powerful Medicines (2004), on prescription drugs. Avorn said in an interview that companies are frequently allowed to submit pivotal studies that lasted only a few months, even for drugs that will be taken for a long time, and to recruit subjects who are not representative of the population likely to take a drug after it is marketed. The agency has also been increasingly willing to grant "accelerated approval" on the basis of a drug's effect on a surrogate marker, such as tumor shrinkage for a cancer drug or reduction of low-density lipoprotein cholesterol for a statin. Federal rules require companies receiving accelerated approval to conduct post-marketing studies that use more definitive outcome measures, but more often than not, companies fail to do so — and the FDA has never responded to such lapses by withdrawing a drug. Of more than 1300 post-marketing studies to which drug companies have committed themselves, 65 percent have not been started.2
David A. Kessler, who was FDA commissioner when this program began, said that PDUFA was meant to provide the FDA with the necessary resources to speed up reviews, not to lower the bar for approval. He added that during his tenure he noticed no lowering of standards. "PDUFA has nothing to do with debate and discussion" within the FDA, he said. "That's leadership. That's the culture of the agency."
Have faster reviews by the FDA contributed to an increase in the number of drugs withdrawn from the market? The answer isn't clear. In the 2002 GAO report, investigators noted an increase in the rate of safety-related drug withdrawals (as expressed by the ratio of the number of drugs withdrawn for safety reasons each year to the number of drugs approved). From 1997 through 2000, the withdrawal rate was 5.34 percent, as compared with 1.96 percent between 1989 and 1992 (pre-PDUFA) and 1.56 percent between 1993 and 1996 (immediately after the enactment of PDUFA). But a recent FDA analysis, which examined drug withdrawals according to year of approval, showed little change after PDUFA was enacted. From 1971 through 1993, the rate of safety-based drug withdrawals was 2.7 percent, as compared with 2.3 percent for drugs approved between January 1, 1994, and April 30, 2004.3
Safety-related withdrawals occur sporadically and for varying reasons — for example, there were five such withdrawals between September 1997 and September 1998, but none in 2003. Recently withdrawn drugs have included one, fenfluramine, that was approved in 1973 after a 75-month review and two others, troglitazone and cerivastatin, that were approved in 1997 after reviews lasting 6 months and 12 months, respectively. Rofecoxib is among the most swiftly approved drugs to be withdrawn for safety reasons. Considered potentially safer for the gastrointestinal tract than older nonsteroidal antiinflammatory drugs, it received a priority six-month review.
Among the most worrisome signs that things are amiss within the agency are reports that FDA scientists have been discouraged by supervisors from raising questions about drug safety and sometimes have been prevented from sharing their concern with FDA advisory committees. In February 2004, medical reviewer Andrew Mosholder was prevented from presenting to an advisory panel his analysis linking the use of antidepressant drugs in teenagers with an increased risk of suicidal behavior; instead, his superiors ordered a second analysis by another FDA scientist, which reached the same conclusion six months later.4 Barbehenn, the former FDA pharmacologist, said that medical officers on her CDER review team were prevented by supervisors from presenting an analysis on alendronate (Fosamax) at an advisory committee meeting. Graham said in an interview that he and other scientists from the Office of Drug Safety sometimes attend advisory-committee meetings but are rarely invited to present data or analyses.
This past summer, faced with supervisors' demands that he amend the conclusions of his Vioxx study, Graham did so, he told the Senate Finance Committee, in order to obtain permission to present the findings at an international conference. "I changed them to a fair degree, and it caused me a great deal of mental anguish," he said. "I did it because I thought if I didn't, there was no way that the data would see the light of day."
Last month, Health and Human Services Secretary Mike Leavitt announced plans to establish a new Drug Safety Oversight Board within the FDA that will draw on outside expertise to review safety issues that arise with regard to approved drugs. Graham's supervisors also permitted him to present findings from an unpublished study to the FDA advisory committee reviewing the safety of cyclooxygenase-2 (COX-2) inhibitors — a decision for which Graham publicly thanked Crawford. These are encouraging signs of change. But some experts have suggested that to achieve impartial scientific assessments of risks, the drug-safety program should be moved out of the FDA into an independent agency. "You can't run a safety system where you appear to be suppressing conflicting opinions, even if these conflicting opinions turn out to be wrong," said Vanderbilt's Wood.
Source Information
Dr. Okie is a contributing editor of the Journal.
References
Office of Inspector General. FDA's review process for new drug applications: a management review. March 2003. (OEI-01-00590.) (Accessed February 24, 2005, at http://www.oig.hhs.gov/oei/reports/oei-01-01-00590.pdf.)
Food and Drug Administration. Report on the performance of drug and biologics firms in conducting postmarketing commitment studies: availability. Fed Regist 2004;69:12162-12164.
Food and Drug Administration. CDER report to the nation: 2003. (Accessed February 24, 2005, at http://www.fda.gov/cder/reports/rtn/2003/rtn2003-3.htm.)
Vendantam S. FDA study confirms antidepressant risks: drugs linked to more suicides among children, unpublished analysis says. Washington Post. August 10, 2004:A6.(Susan Okie, M.D.)