Cox-2s safety summit sees FDA scientist censored

Click to follow
The Independent Online

David Graham, the high-profile scientist at America's Food and Drug Administration, has been banned from presenting new information about a controversial type of painkiller at a three-day summit on their safety starting today.

David Graham, the high-profile scientist at America's Food and Drug Administration, has been banned from presenting new information about a controversial type of painkiller at a three-day summit on their safety starting today.

Dr Graham, who works in the FDA's office of drug safety, said yesterday that he had been "threatened with being called insubordinate" by the FDA when he said he wanted to include the findings of an unpublished study he has completed as part of his testimony on Cox-2 inhibitors.

The FDA has organised the series of hearings, which will include evidence from drug makers and scientists, in Washington in response to mounting concerns about Cox-2s and to allegations that the agency has been too slow to deal with possible dangers associated with the drugs. Last September, the US pharmaceuticals giant Merck withdrew Vioxx after a study showed the blockbuster increased the risk of heart disease five-fold in patients on a high dose.

In an attempt to head off further criticism of the FDA, the US Secretary of Health and Human Services, Mike Leavitt, announced a new independent Drug Safety Oversight Board would be created within the FDA to monitor medicines once they are on the market and to update physicians and patients with emerging information on risks and benefits.

Dr Graham, who has accused the FDA of ignoring worrying evidence about Vioxx when it was still on the market, said he wanted to include in his testimony tomorrow details of a study he completed with Gurkirpal Singh of Stanford University School of Medicine.

This study "talks about the risks of Cox-2 agents and other drugs that are widely used," he said. "We think the findings are important. The committee might not think so, but it should have the opportunity to consider it," Dr Graham said.

As well as Vioxx, drugs which will be scrutinised at the meeting are Celebrex and Bextra, two Cox-2s made by Pfizer.

The hearings will be monitored closely by the UK's GlaxoSmithKline, which has stopped work on its own Cox-2 pending clarification of the requirements for large-scale human trials.

The FDA is not expected to ban Cox-2s outright, and GSK is unlikely to abandon work on its drug, codenamed 381. The company says it had always expected trials involving thousands of patients would be required and that it had budgeted for increased expenditure since Vioxx was withdrawn. It has already done extra work on the existing data, and says the effect Vioxx appears to have on blood pressure is not found with 381.

GSK believes head-to-head trials against Celebrex, the market leader have proved that 381 is a superior drug, and the outcome of this week's FDA hearings and more talks with the agency this spring will determine how long it takes to bring the product to the US market.

The FDA panel is thought likely to demand that manufacturers increase the level of warning on bottles, and perhaps carry out trials into the long-term impact of the drugs when they are on the market. It could order that trials are designed to generate more data on high-risk patients.

Relations between Dr Graham and his employer have been strained since he spoke out against the way the FDA handled Vioxx, which he said caused heart problems in up to 140,000 patients.

Last November, the FDA publicly repudiated comments by Dr Graham naming five drugs, including AstraZeneca's cholesterol lowering pill Crestor and GSK's asthma drug Serevent, that he thought could be too dangerous for use. The agency argued that it has struck the right balance of benefits and risks when assessing new drugs for launch, but it has been accused of being too close to the drug industry and has faced political calls to split its safety monitoring functions into an agency separate from the drug approvals process.

Comments