FDA Advisory Panel Nixes New Cox-2

A Food and Drug Administration advisory panel voted 20-1 today against approving Merck's new arthritis pain pill, known as Arcoxia. After listening to the company's presentation, the FDA reviewer's analysis, and a devastating critique offered by FDA safety officer David Graham, one could only wonder why Merck brought this Vioxx clone before the committee.

It also makes one wonder why over 60 European countries have approved the drug, which appears to have all the same heart attack risks of other Cox-2 inhibitors. The evidence that the drug reduced the gastrointestinal side effects of other pain pills like ibuprofen or naproxen (sold over-the-counter as Aleve) was marginal at best, and was clearly outweighed by the more serious cardiovascular side effects.

And, the bottom line was that the drug provided no more additional pain relief than any other non-steroidal anti-inflammatory drug (NSAIDs). I could not believe this morning's radio report by Helen Palmer on Marketplace. She called Arcoxia "super aspirin" and quoted a Merck p.r. flack claiming the drug was superior to other NSAIDs. The evidence offered at the hearing today was incontrovertible on that point: it provided no more pain relief than any other drug.

Despite the presence of a number of physicians on the committee who clearly would like to see more and better pain relievers for their arthritis patients, Merck stepped forward with an extremely weak set of data. They tested the drug in a staggering 34,000-plus patients, half of whom took diclofenac (sold as Voltaren or Cataflam), which has very low market penetration in the U.S. What it does have is a heart attack risk profile that is very similar to other Cox-2 inhibitors.

"What is the value of comparing it to diclofenac, which has an elevated cardiovascular risk?" asked Graham near the conclusion of his devastating critique. "It has no value."

According to Graham, if Arcoxia had been compared to naproxen, it would have shown cardiovascular risk nearly three times higher than the over-the-counter medication.

How about the supposed G.I. benefits of Arcoxia? For the first time, one of the Cox-2s appears to have actually shown that there was some reduction in the gastrointestinal side effects found in most NSAIDs (it only effects around 2 percent of the people who take them). In Merck's massive study, the number of GI adverse events (those needing physician and medical intervention) declined to 176 in the Arcoxia group compared to 246 in the diclofenac group (a reduction from 1.4 to 1.0 percent of the overall population taking the drugs). Virtually all the reductions came in non-life threatening bleeding and ulcers. More serious conditions like G.I. tract perforations and obstructions were about the same in the two groups.

On the other hand, both groups had over 300 serious cardiovascular complications, including more than 100 heart attacks. Had either group (Arcoxia or diclofenac) been taking naproxen, they would have suffered nearly 60 percent fewer cardiovascular "events." David Egilman, a physician at Brown University who works as a plaintiffs expert in Vioxx litigation and testified during the public portion of the meeting, called the Merck trial "unethical."

There was one personally gratifying moment during the hearing. David Graham began his presentation by debunking claims, often carried in the media, that NSAIDs lead to 16,500 deaths a year from G.I. side effects. He estimated that G.I. complications in the U.S. cost about 4,714 lives a year, and many of those have nothing to do with NSAIDs.

After his presentation, James Fries, a gastroenterologist from Stanford University and a member of the advisory panel, grabbed his microphone. "I've been ashamed of that number ever since it came out," he said. A former colleague at Stanford generated the 16,500 deaths number in the late 1990s and it became a major talking point for the marketing departments at Pfizer and Merck when they began pushing Celebrex and Vioxx.

Caught in the hallway during a meeting recess, Fries explained that the original estimate had been based on 1992 data that included a high number of rheumatoid arthritis sufferers. These are people who take lots of pain pills and are more likely to suffer G.I. side effects.

Moreover, he explained, the incidence of G.I. complications from NSAID use began declining almost immediately after that data was compiled due to better painkillers and better acid indigestion fighters (like Zantac and later Prilosec). "The incidence rate has been cut by a third. The death rate is at most 2,500 now and even that may be too high," he said.

I've been making this point for years. Now there will be confirmation in an FDA hearing transcript. Hopefully, we've heard the last press claim that NSAIDs cause "16,500 deaths annually."

Posted by gooznews at April 12, 2007 06:06 PM