FDA panel discusses risk in Cox-2 inhibitors
Suggest improvements in translating risk-to-benefit profile
By Jeanie Kennedy
The Food and Drug Administration’s (FDA) Arthritis Advisory Committee and the Drug Safety and Risk Management Advisory Committee (Panel) held a joint meeting on February 16-18, 2005, to discuss cardiovascular risks for non-selective and selective Cox- 2 non-steroidal anti-inflammatory drugs (NSAIDs).
The Panel found there to be a general class effect of increased thrombosis in the Cox-2 inhibitors already approved for marketing in the United States. These drugs include: rofecoxib (Vioxx), celecoxib (Celebrex), and valdecoxib (Bextra). The Panel attributed different mechanisms of the drug class’s actions to differences in patient outcomes, and deemed the entire class as worthy of future study. Additionally, Panel members agreed that higher doses of Cox-2 inhibitors provide more risk of thrombotic events, with the hazard ratio increasing in direct proportion to the length of consumption of the Cox-2 inhibitors.
The use of Cox-2 inhibitors postoperatively in patients undergoing orthopaedic surgical procedures known to be associated with increased risk of thromboembolic disease was not discussed extensively and no formal recommendations were made by the Panel.
Hierarchy of evidence
During the meeting, various stakeholders presented information from multiple case-controlled series and prospective trials; the Panel placed a greater emphasis on outcomes from placebo-controlled trials. The Panel found an increase in myocardial infarction and stroke in five placebo-controlled trials with the three structurally distinct Cox-2 inhibitors.
Although the Alzheimer’s Disease and Prevention Trial (ADAPT) safety data did not warrant suspension of celecoxib treatments, the trial was halted on December 20, 2004, for operational problems due to a 40 percent patient dropout rate after the recall of rofecoxib. Safety data will be published in a peer-reviewed journal article at a future date. The data did show a slight increase in cardiovascular risk with the use of naproxen.
The Panel also did not find naproxen to be cardioprotective, but did laud the drug for its use in arthritis trials, finding the use of placebo to be untenable. Naproxen, well studied in millions of patients, was encouraged for use in future trials in comparisons of efficacy and cardiovascular risks for NSAIDs.
The Panel also noted the process of translating information to physicians about pharmaceutical risks should be improved. Risk-to-benefit balance for each drug must be available and clearly understood for patients and physicians to engage in informed decision-making.
The Panel deliberated on the data from the Therapeutic Arthritis and Gastrointestinal Event Trial, ADAPT, and Vioxx Gastrointestinal Outcomes Research studies. David Graham, MD, MPH, medical officer with the Center for Drug Evaluation and Research (CDER)/FDA, commented that patients in randomized clinical trials tend to be healthier than those in the general U.S. population.
Panel members also suggested that observational studies should be used to determine research questions, and randomized clinical trials should follow. Observational studies provide data on the actual use of medication in the U.S. population. Some Panel members recommended the ongoing study of Cox-2 inhibitors for at least three to four years given that this class of drugs is already widely used in the U.S. population. Panel members also commented that the quality of the clinical trials could be improved.
The Panel voted on whether the risk/benefit profile supported continued marketing of Cox-2 inhibitors. Voting results narrowly supported the continued marketing of rofecoxib (17-15) and valdecoxib (17-13 with two abstentions). Celecoxib received a vote of 31-1 for continued marketing.
Most Panel members advocated the use of “black box” warnings on these drugs to communicate potential cardiovascular risks. Other recommended options included: restricting use to a small patient population; a ban on all direct- to-consumer advertising for the class; creation of patient guides; more detailed package inserts; increased postmarketing surveillance; and more clinical studies. “Black box” warnings effectively deter manufacturers’ from producing direct-to-consumer advertising.
Panel members also recommended empowering the FDA to enforce postmarketing study commitments and postmarket surveillance. Continued postmarket surveillance is essential to capture problems in the general population.
The Panel did not recommend the use of Cox-2 inhibitors in patients with ischemic cardiovascular risk, and recommended that physicians should thoroughly discuss risk/benefit ratios with their patients before prescribing the drugs.
Panel recommendations regarding the NSAID drug class include a warning on the labels of older NSAIDs, including naproxen and ibuprofen, to advise patients of a possible cardiac risk.
FDA Panel recommendations are nonbinding. Nonetheless, the FDA usually accepts the conclusions and recommendations of its expert advisory committees. Steven Galson, MD, MPH, acting director of CDER, promised to act quickly on these recommendations.
Jeanie Kennedy is manager, biomedical research and regulation, in the AAOS department of research and scientific affairs. She can be reached at firstname.lastname@example.org