Vioxx blamed for 300 Australian deaths

A Royal Adelaide Hospital cardiovascular disease expert has told ABC-TV's current affairs program Four Corners that Merck & Co's anti-inflammatory drug Vioxx may have contributed to the deaths of many as 300 Australians from heart attack or stroke.

The program, due to be screened tonight, also airs claims by US lawyers who say they have obtained internal memos and other documents from Merck showing that the international drug giant knew in the 1990s that Vioxx was potentially dangerous -- a claim Merck has denied.

The estimate of 300 Australian deaths due to the novel cyclooxygenase-2 (COX-2) receptor inhibitor was made by Royal Adelaide Hospital's Prof Les Cleland.

The US Food and Drug Administration ordered Merck to take Vioxx off the global market last September, after medical evidence linked its best-selling anti-inflammatory and pain reliever to more than 100,000 heart attacks and strokes in the US.

The risk was uncovered during a trial in patients with colorectal polyps, who are at increased risk of colorectal cancer. The trial compared Vioxx with the standard anti-inflammatory drug naproxen.

Cleland told Four Corners that about 30 per cent of 1000 Australian deaths from heart attack and stroke may be linked to Vioxx therapy. According to the Four Corners investigation, there is still concern about other COX-2 inhibitors, such as Pfizer's Celebrex.

Extrapolations 'risky'

But another leading Australian expert on heart disease, Prof Henry Krum, director of the NHMRC Centre of Clinical Research Excellence in Therapeutics at Monash University, told Australian Biotechnology News there was a risk in making population-wide extrapolations from a limited number of adverse events.

"I'm still to be convinced that there is inherent increase in risk with all drugs in the COX-2 inhibitor class," Krum said. "There's no doubt it's there for rofecoxib [Vioxx], especially in colorectal polyp patients, who were the death knell for the drug. But with the other drugs, like celecoxib [Celebrex] and valdecoxib [Pfizer's Bextra] the epidemiology suggests there is no signal. Some researchers are saying that if rocefoxib didn't exist, we wouldn't be having this controversy.

"Just to make things more interesting, there is some data suggesting that COX-2 inhibitors may actually benefit patients with cardiovascular disease."

Krum said all the studies to date had been conducted in patients at low risk of cardiovascular disease. High-risk patients had been excluded because of the potentially confounding effects of other drugs they were taking for their condition.

He described the supposed risk of other COX-2 inhibitors to cardiovascular-disease patients as "a complete extrapolation from the data". Usually, regulatory authorities like the FDA and Australia's Therapeutic Goods Administration based their approvals or non-approvals on trial data from a particular cohort.

"All these studies of COX-2 inhibitors were conducted in patients with a low risk of cardiovascular disease," he said. "The COX-2 inhibitors have never been studied in high-risk patients.

"The COX-1 inhibitors have been lumped together as a class, but they have different pharmacokinetic characteristics, quite different structures, and different metabolism."

Biological rationale

Prof Mark Nelson, head of the Discipline of General Practice at the University of Tasmania, was one of the authors of an editorial on the COX-2 inhibitor affair in the Medical Journal of Australia last year, which said there was a biological rationale that might have predicted the problem with Vioxx -- drugs that selectively inhibit the COX-2 but not the COX-1 receptor could promote clotting by inducing over-expression of prostacyclin, which has a markedly enhanced action in atherosclerosis.

Nelson said a possible reason why rofecoxib had been identified as a higher risk drug in colorectal polyp patients was that it was actually inferior in this specific application to the classic non-steroidal anti-inflammatory (NSAID) drug naproxen, with which it was being compared.

"Classic NSAIDs like aspirin bind irreversibly to platelets to prevent them aggregating and forming blood clots," he said. "We have reason to suspect the COX-2 inhibitors aren't doing that, so when they go head-to-head with a classic NSAID, they may be inferior.

"But the main point of all this is that any new drug group has the potential to cause long-term, adverse effects because of the paucity of data at the time of their public release.

"Trials are designed to answer a specific clinical question. If a blood-pressure reducing drug lowers blood pressure efficiently, without significant side effects in the patient cohort, it will be approved -- as long as it works as well as rival drugs, and reduces the risk of stroke.

"But when you introduce a new drug, with a different mechanism of bioactivity, you will have the least understanding of its long-term effects. The classic example is that the aged take most medications, yet they are the least likely to appear in clinical trials, because researchers want to eliminate confounding factors."

The result, Nelson said was that trials tended to involve 'purer' patient cohorts that might not reflect of the diversity of responses in the general population.

One way around the problem, Nelson suggested, was to require drug companies to put some of their profits from new drugs aside for randomised trials in groups where there was some indication of potentially adverse effects, even if such effects did not reach the level of statistical significance in trials.

"You might be able to identify a particular problem there and then," Nelson said. "The trial results mightn't be significant, or suggest causality. They might be a false positive -- but at least they would provide a starting point, a red flag, for monitoring for adverse long-term effects."