Oblique reference to MLNM in Barron's:
online.barrons.com
Last week, the Food and Drug Administration put out guidelines for companies that use genetic tests in drug development. Such "pharmacogenetic" tests examine genetic differences in how patients respond to a drug. These differences can render drugs ineffective or unsafe in patients who are short of certain enzymes, for example. Some drug companies had feared the agency would use genetic profiles to restrict the market size of a potential blockbuster drug -- by limiting approval to just the subgroup of the general population that best metabolizes the drug.
The new guidelines show the industry that the FDA has no plans to genetically redline drugs. Indeed, the agency hopes pharmacogenetic testing will spread from the drug development lab to the clinic, where doctors could choose drugs and doses based on a patient's genetic traits. To encourage companies to develop products for such "personalized medicine," the FDA will soon release a concept paper that discusses how companies might develop a pharmacogenetic test that's paired with a new drug.
Widespread practice of personalized medicine would be a windfall for Affymetrix (AFFX), the Santa Clara, Calif.-based maker of the DNA microarray chips used for most pharmacogenetic work. In January, the FDA approved a test called the AmpliChip CYP450, developed by Affymetrix and Roche Diagnostics to screen for variations in two genes that affect how patients metabolize some 25% of commonly prescribed drugs. Use of the test would prevent dangerously wrong doses of codeine or blood-pressure drugs, says Heino von Prondzynski, who heads Roche Diagnostics. The Roche executive had everyone in his family tested for their CYP450 profile. He discovered that his two-year-old grandson will metabolize certain medicines too slowly.
A big clinical market for DNA microarrays might even justify the generous valuation of Affymetrix shares. On a 15% rise in sales last year, to $346 million, profits tripled to 74 cents a share. At about 45, the stock trades at over 60 times last year's earnings and 35 times the estimate for 2006. But Roche and Affymetrix have other AmpliChip tests under development. In the next year, they hope to seek approval for a test to flag the P53 genetic sequence -- a marker of especially aggressive tumors. Also in the pipeline is an array that looks at about 600 genes involved in leukemia, the blood cancer that occurs in dozens of varieties -- many needing different treatments.
While pharmacogenetic tests are a powerful new tool, a person's genetic profile is usually just one a number of factors affecting her drug response. So instead of restricting sales, the FDA has required disclosure of genetically related side effects in the package inserts for drugs. When the leukemia drug 6MP proved to be highly toxic in children with a particular genetic profile, the FDA consulted experts and ultimately did not make genetic testing a requirement for the drug's continued sale.
If the FDA has shown a light touch in handling pharmacogenetic know-how, the technology may yet impact drug sales if the tests become popular with healthcare payers. AstraZeneca's cancer drug Iressa only slowed tumor growth for those 10% of patients with a particular genetic mutation. "What's the benefit for the other 90%?" wonders Christopher J. Webster, the director of regulatory strategy for the Cambridge, Mass., biotech firm Millennium Pharmaceuticals.
But despite the potential for using gene profiles to control costs, insurers have held off. Webster knows of insurers who reimburse for the Genentech breast cancer drug Herceptin even if the tumors lack the drug's genetic target.
For its part, Millennium makes heavy use of DNA microarrays. The company scours the genomes of patients afflicted with rheumatoid arthritis, looking for genetic markers of disease. It's developing a cancer drug that would be marketed with a matching genetic test. Most other sophisticated drug firms are keen on pharmacogenetics, too, and the FDA wants to be a help and not a hindrance. |
