Raising the Safety Bar — The FDA's Coxib Meeting
http://www.100md.com
《新英格兰医药杂志》
After sitting through three days of testimony and debate last month, a panel of experts advising the Food and Drug Administration (FDA) rendered an ambivalent verdict on the future of the cyclooxygenase-2 (COX-2) inhibitors. On the one hand, the 32 members of the advisory committee agreed that these drugs for pain and arthritis pose serious hazards. They voted unanimously that the available evidence on the three approved "coxibs" — rofecoxib, celecoxib, and valdecoxib — supports the conclusion that each of the drugs significantly increases users' risk of cardiovascular events, including heart attacks and strokes. Nonetheless, the committee also voted, by a different margin in each case, in favor of allowing the medications to continue to be sold in the United States, although they recommended placing black-box warnings on their labels and instituting other measures to severely restrict their use. (Merck, the maker of rofecoxib , voluntarily withdrew its drug from the market last fall, but the company now indicates that it may soon bring the product back.)
The votes in favor of continued marketing of the drugs came despite a warning by the committee chairman, pharmacologist Alastair J.J. Wood of Vanderbilt University Medical School, that the evidence of cardiovascular risk "is a far larger safety signal than we have seen for any of the other drugs withdrawn for safety reasons." Whereas drugs have usually been withdrawn because they have been found to increase the risk of a rare condition, such as hepatic failure, any widely used drug that increases the incidence of cardiovascular disease, the leading cause of death in the United States, may well have a far bigger effect on the health of the population. "This is probably one of the first times that we and the FDA have had to deal with a drug that causes a substantial increase in a common problem," Wood said.
The recommendations of FDA advisory committees are not binding, but the agency usually follows them. During the discussion, some committee members — especially the rheumatologists among them — argued that the coxibs can be clinically valuable for certain patients with osteoarthritis or rheumatoid arthritis who have not had adequate pain relief from other nonsteroidal antiinflammatory drugs (NSAIDs) or who are unable to tolerate the gastrointestinal side effects of other drugs. (When the committee invited public comment, several people with severe arthritis or chronic pain offered testimonials regarding the benefit they had received from coxibs, although Wood later noted that the drugs have never been proved superior to other NSAIDs for pain relief in clinical trials, and only rofecoxib has been shown to reduce the incidence of gastroduodenal ulcers and bleeding.) Some committee members also voiced the concern that some of the older NSAIDs — especially drugs such as diclofenac or meloxicam that inhibit COX-2 more than cyclooxygenase-1 (COX-1) — may turn out to pose similar cardiovascular risks, although their adverse effects have not been as thoroughly studied. "We will, by our actions here, cause a shift in prescribing practices," predicted Steven Nissen, a cardiologist at the Cleveland Clinic. "That shift should, to the best of our ability, be a shift toward greater safety."
Indeed, a recurring theme of the meeting was the need to conduct additional studies of the long-term safety of virtually all NSAIDs, focusing particularly on their effects on cardiovascular risk and blood pressure. There are far more data from controlled trials involving rofecoxib and celecoxib than on the older NSAIDs, and almost no clinical trials have studied the safety of these older drugs in comparison with placebos. Many of the safety data on the older NSAIDs come from studies in which one such drug (most often naproxen, diclofenac, or ibuprofen) was used as the standard of comparison to assess the efficacy of a coxib in patients with arthritis. Because a placebo cannot ethically be included in arthritis studies, we have no way of estimating the relative risk of an adverse event such as a heart attack among patients taking a particular drug as compared with patients taking no drug at all. This limitation created problems in interpreting the results of the Vioxx Gastrointestinal Outcomes Research (VIGOR) trial, a Merck-sponsored study submitted to the FDA in 2000 in which rofecoxib was compared with naproxen; when the data revealed an increased frequency of cardiovascular events in the rofecoxib group, the study's authors contended that the findings reflected a protective effect of naproxen, rather than a deleterious effect of rofecoxib.1
Thus, the committee members found themselves faced with sufficient data to convict the coxibs of increasing cardiovascular risk but not enough, so far, to convict any of the older NSAIDs of similar hazards. "At the same time, I am reluctant to give them a get-out-of-jail-free card," epidemiologist Curt D. Furberg of Wake Forest University commented. Furberg made an exception, however, for naproxen, which has been studied extensively as a comparison drug, sometimes in trials that included a placebo, and appears to have either a neutral or a slightly beneficial effect on cardiovascular risk. "Naproxen looks good," he acknowledged.
Robert J. Temple, associate director for medical policy at the FDA's Center for Drug Evaluation and Research, argued that a large, multigroup trial should be conducted to assess the short-term and long-term safety of the entire class of NSAIDs. Participants in such a study should be patients with osteoarthritis, rheumatoid arthritis, or chronic pain, stratified according to their baseline level of cardiovascular risk. Temple proposed that such a trial compare ibuprofen, naproxen, diclofenac, and celecoxib groups with a "control" group that would receive either full-dose aspirin plus a proton-pump inhibitor or acetaminophen plus codeine. The end points should include death from cardiovascular disease, stroke, acute myocardial infarction, and bleeding; blood pressure would be monitored and would be treated as necessary. Patients at high risk for coronary artery disease (except for those in the full-dose–aspirin group) would also receive low-dose aspirin. Temple estimated that at least 1000 participants would be needed in each group and that such a trial would take several years to complete. Who would fund such a trial is unclear.
The committee's decisions regarding the individual COX-2 inhibitors reflected the members' sense that the degree of cardiovascular risk associated with celecoxib appeared smaller than that with rofecoxib and valdecoxib. They voted 31 to 1 in favor of allowing celecoxib to remain on the market. In contrast, the vote on rofecoxib was almost evenly split, with 17 voting in favor of allowing it to be marketed and 15 opposed. The data on rofecoxib's cardiovascular risk "are very compelling," noted Wood. The database from clinical trials of valdecoxib was much smaller than those for the other two drugs, but there was a strong risk signal from a study of the drug in recipients of coronary bypass grafts.2 The panel voted 17 to 13 in favor of keeping it on the market, with 2 members abstaining. It was later reported, however, that 10 members of the committee, a subgroup who voted overwhelmingly to permit continued use of the coxibs, had financial ties to the manufacturers of the drugs in question.3
At times during the meeting, Wood sharply criticized officials from both Merck and Pfizer for omitting from their presentations information regarding serious adverse events associated with their drugs. For example, he asked why a Pfizer presenter had not included data on the increased risk of congestive heart failure among celecoxib users that was found in a study released the previous day on the Journal's Web site.4 The decision not to report the finding "just doesn't pass the laugh test," Wood said.
In discussing what restrictions should be placed on the distribution of the coxibs, almost all committee members said that manufacturers should be prohibited from advertising the products directly to consumers. Studies indicate that such advertising makes it much more likely that a drug will be used by patients who do not have the indications for which it is approved. Forbidding such advertising "would prevent more serious adverse events than anything else we could do, other than taking the drugs off the market," Furberg said.
"We don't think we can ban direct-to-consumer advertising," responded the FDA's Temple. However, requiring a black-box warning on the coxibs' labels would prevent manufacturers from running "reminder ads" that simply mention a product and urge patients to ask their doctors about it.
During the discussion, many committee members also indicated that they would be reluctant to recommend approval for new COX-2 inhibitors, or to favor allowing additional NSAIDs to be sold without a prescription, until more safety data are collected on the entire class of drugs. They noted that, in the future, new NSAIDs being considered for FDA approval are likely to be held to a higher safety standard, including a requirement for clinical studies lasting at least a year and preferably two to three years.
"We've got a lot of evidence that several drugs in this class can elevate the risk" of cardiovascular disease, said the Cleveland Clinic's Nissen. "Now that we know that, we know where to set the bar for these drugs, and it's got to be set pretty high."
Source Information
Dr. Okie is a contributing editor of the Journal.
References
Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:1520-1528.
Nussmeier NA, Whelton AA, Brown MT, et al. Complications of the COX-2 inhibitors parecoxib and valdecoxib after cardiac surgery. N Engl J Med 2005;352:1081-1091.
Harris G, Berenson A. 10 Voters on panel backing pain pills had industry ties. New York Times. February 25, 2005:A1.
Solomon SD, McMurray JJV, Pfeffer MA, et al. Cardiovascular risk associated with celecoxib use in a large, randomized clinical trial for colorectal adenoma prevention. N Engl J Med 2005;352:1071-1080.(Susan Okie, M.D.)
The votes in favor of continued marketing of the drugs came despite a warning by the committee chairman, pharmacologist Alastair J.J. Wood of Vanderbilt University Medical School, that the evidence of cardiovascular risk "is a far larger safety signal than we have seen for any of the other drugs withdrawn for safety reasons." Whereas drugs have usually been withdrawn because they have been found to increase the risk of a rare condition, such as hepatic failure, any widely used drug that increases the incidence of cardiovascular disease, the leading cause of death in the United States, may well have a far bigger effect on the health of the population. "This is probably one of the first times that we and the FDA have had to deal with a drug that causes a substantial increase in a common problem," Wood said.
The recommendations of FDA advisory committees are not binding, but the agency usually follows them. During the discussion, some committee members — especially the rheumatologists among them — argued that the coxibs can be clinically valuable for certain patients with osteoarthritis or rheumatoid arthritis who have not had adequate pain relief from other nonsteroidal antiinflammatory drugs (NSAIDs) or who are unable to tolerate the gastrointestinal side effects of other drugs. (When the committee invited public comment, several people with severe arthritis or chronic pain offered testimonials regarding the benefit they had received from coxibs, although Wood later noted that the drugs have never been proved superior to other NSAIDs for pain relief in clinical trials, and only rofecoxib has been shown to reduce the incidence of gastroduodenal ulcers and bleeding.) Some committee members also voiced the concern that some of the older NSAIDs — especially drugs such as diclofenac or meloxicam that inhibit COX-2 more than cyclooxygenase-1 (COX-1) — may turn out to pose similar cardiovascular risks, although their adverse effects have not been as thoroughly studied. "We will, by our actions here, cause a shift in prescribing practices," predicted Steven Nissen, a cardiologist at the Cleveland Clinic. "That shift should, to the best of our ability, be a shift toward greater safety."
Indeed, a recurring theme of the meeting was the need to conduct additional studies of the long-term safety of virtually all NSAIDs, focusing particularly on their effects on cardiovascular risk and blood pressure. There are far more data from controlled trials involving rofecoxib and celecoxib than on the older NSAIDs, and almost no clinical trials have studied the safety of these older drugs in comparison with placebos. Many of the safety data on the older NSAIDs come from studies in which one such drug (most often naproxen, diclofenac, or ibuprofen) was used as the standard of comparison to assess the efficacy of a coxib in patients with arthritis. Because a placebo cannot ethically be included in arthritis studies, we have no way of estimating the relative risk of an adverse event such as a heart attack among patients taking a particular drug as compared with patients taking no drug at all. This limitation created problems in interpreting the results of the Vioxx Gastrointestinal Outcomes Research (VIGOR) trial, a Merck-sponsored study submitted to the FDA in 2000 in which rofecoxib was compared with naproxen; when the data revealed an increased frequency of cardiovascular events in the rofecoxib group, the study's authors contended that the findings reflected a protective effect of naproxen, rather than a deleterious effect of rofecoxib.1
Thus, the committee members found themselves faced with sufficient data to convict the coxibs of increasing cardiovascular risk but not enough, so far, to convict any of the older NSAIDs of similar hazards. "At the same time, I am reluctant to give them a get-out-of-jail-free card," epidemiologist Curt D. Furberg of Wake Forest University commented. Furberg made an exception, however, for naproxen, which has been studied extensively as a comparison drug, sometimes in trials that included a placebo, and appears to have either a neutral or a slightly beneficial effect on cardiovascular risk. "Naproxen looks good," he acknowledged.
Robert J. Temple, associate director for medical policy at the FDA's Center for Drug Evaluation and Research, argued that a large, multigroup trial should be conducted to assess the short-term and long-term safety of the entire class of NSAIDs. Participants in such a study should be patients with osteoarthritis, rheumatoid arthritis, or chronic pain, stratified according to their baseline level of cardiovascular risk. Temple proposed that such a trial compare ibuprofen, naproxen, diclofenac, and celecoxib groups with a "control" group that would receive either full-dose aspirin plus a proton-pump inhibitor or acetaminophen plus codeine. The end points should include death from cardiovascular disease, stroke, acute myocardial infarction, and bleeding; blood pressure would be monitored and would be treated as necessary. Patients at high risk for coronary artery disease (except for those in the full-dose–aspirin group) would also receive low-dose aspirin. Temple estimated that at least 1000 participants would be needed in each group and that such a trial would take several years to complete. Who would fund such a trial is unclear.
The committee's decisions regarding the individual COX-2 inhibitors reflected the members' sense that the degree of cardiovascular risk associated with celecoxib appeared smaller than that with rofecoxib and valdecoxib. They voted 31 to 1 in favor of allowing celecoxib to remain on the market. In contrast, the vote on rofecoxib was almost evenly split, with 17 voting in favor of allowing it to be marketed and 15 opposed. The data on rofecoxib's cardiovascular risk "are very compelling," noted Wood. The database from clinical trials of valdecoxib was much smaller than those for the other two drugs, but there was a strong risk signal from a study of the drug in recipients of coronary bypass grafts.2 The panel voted 17 to 13 in favor of keeping it on the market, with 2 members abstaining. It was later reported, however, that 10 members of the committee, a subgroup who voted overwhelmingly to permit continued use of the coxibs, had financial ties to the manufacturers of the drugs in question.3
At times during the meeting, Wood sharply criticized officials from both Merck and Pfizer for omitting from their presentations information regarding serious adverse events associated with their drugs. For example, he asked why a Pfizer presenter had not included data on the increased risk of congestive heart failure among celecoxib users that was found in a study released the previous day on the Journal's Web site.4 The decision not to report the finding "just doesn't pass the laugh test," Wood said.
In discussing what restrictions should be placed on the distribution of the coxibs, almost all committee members said that manufacturers should be prohibited from advertising the products directly to consumers. Studies indicate that such advertising makes it much more likely that a drug will be used by patients who do not have the indications for which it is approved. Forbidding such advertising "would prevent more serious adverse events than anything else we could do, other than taking the drugs off the market," Furberg said.
"We don't think we can ban direct-to-consumer advertising," responded the FDA's Temple. However, requiring a black-box warning on the coxibs' labels would prevent manufacturers from running "reminder ads" that simply mention a product and urge patients to ask their doctors about it.
During the discussion, many committee members also indicated that they would be reluctant to recommend approval for new COX-2 inhibitors, or to favor allowing additional NSAIDs to be sold without a prescription, until more safety data are collected on the entire class of drugs. They noted that, in the future, new NSAIDs being considered for FDA approval are likely to be held to a higher safety standard, including a requirement for clinical studies lasting at least a year and preferably two to three years.
"We've got a lot of evidence that several drugs in this class can elevate the risk" of cardiovascular disease, said the Cleveland Clinic's Nissen. "Now that we know that, we know where to set the bar for these drugs, and it's got to be set pretty high."
Source Information
Dr. Okie is a contributing editor of the Journal.
References
Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:1520-1528.
Nussmeier NA, Whelton AA, Brown MT, et al. Complications of the COX-2 inhibitors parecoxib and valdecoxib after cardiac surgery. N Engl J Med 2005;352:1081-1091.
Harris G, Berenson A. 10 Voters on panel backing pain pills had industry ties. New York Times. February 25, 2005:A1.
Solomon SD, McMurray JJV, Pfeffer MA, et al. Cardiovascular risk associated with celecoxib use in a large, randomized clinical trial for colorectal adenoma prevention. N Engl J Med 2005;352:1071-1080.(Susan Okie, M.D.)