If drugs could be withheld from patients who have a genetic risk for serious side effects, it could not only protect the patients but might help manufacturers get their drugs approved or avoid having to remove them from the market.
"We have to remove drugs that are clearly offering patients benefits because a handful are getting these severe adverse reactions," said Duncan McHale, a researcher for Pfizer and co-chairman of the consortium's scientific management committee.
The consortium will start by trying to find genetic predictors of two side effects serious liver toxicity and Stevens-Johnson syndrome, a rare but potentially fatal blistering of the skin. Both effects are associated with numerous drugs.
McHale said the side effects were rare enough that no single company had enough cases and enough patient DNA samples to find genetic risk factors. "The way to do the best science is for all to work together to get as many cases as possible," he said.
Arthur Holden, the chief executive of the consortium, said the companies would contribute guidance and money and the results will be put into the public domain, with none of the corporate sponsors having early access or being able to patent the findings, he said. Once the results are made public, any company or laboratory would be able to develop and sell genetic tests to predict side effects.
The new Serious Adverse Events group would rely largely on DNA samples that European academic researchers had already collected from people who suffered the severe liver toxicity or Stevens-Johnson syndrome after taking drugs.
Using DNA analysis chips, researchers will examine those patients' DNA at one million different spots, comparing it with the DNA of control subjects, to look for variations between the groups. The DNA of about 2,000 people will be analyzed, with results available in a year or two, Holden said.
The raw results would require further studies to show if the markers really do predict who might suffer the side effects. It is possible that no such markers will be validated.
Even if relevant genes are identified, it might not be economically possible to test all patients to screen out side effects that show up in as few as one in 100,000 people. Such economic factors are one reason that pharmacogenetics, the practice of tailoring drug therapy to patients using genetic tests, has been slow to catch on.
The consortium trend is seen as a possible way out. In addition to Pfizer, the companies involved in the new venture are Abbott Laboratories, GlaxoSmithKline, Johnson & Johnson, Roche, Sanofi-Aventis and Wyeth.
The Food and Drugs Administration (FDA) has encouraged such groups, even helping to start some of them though a program called the Critical Path Initiative, which aims to speed drug development.