Mar 23 2005
Many experts believe a milestone has been reached with the issuing of guidelines for pharmaceutical companies to encourage them to gather and submit information about how genetic variations affect the way people respond to their drugs.
In the past there has been some reluctance on the part of drug companies do this because of concern that the government might use the information to limit the market for their drugs.
Drug manufacturers have also worried that the FDA might block approval if, for example, a drug was shown in a laboratory test to activate a gene that might be involved in cancer. The guidelines are a significant step towards the development of so-called personalized medicine, in which drugs would be tailored to individuals based on their genetic profiles.
The FDA in order to reassure drug companies, says that the majority of the genetic information will not be considered in regulatory decisions, but will be used to help the development of a new science, pharmacogenomics or pharmacogenetics, which aims to tie drug response to genetic variations. Janet Woodcock, the FDA acting deputy commissioner for operations, says that was not previously being done either because drug companies did not understand the regulatory status or it just was not being submitted to the FDA She says it is vitally important that medicine moves into this paradigm.
Edward Abrahams, executive director of the Personalized Medicine Coalition, a Washington group of companies, academic institutions and others that aim to promote the new field, says it demonstrates the government is now an official advocate for personalized medicine. The new science of pharmacogenomics is still in its infancy and though it is common knowledge that a particular drug might work for some people but not for others, or might cause side effects in some people but not in others, there has been little testing of patients for genetic variations that could predict such reactions, which would allow drugs to be given to appropriate patients and reduce side effects.
A good example of a case in point is the drug Herceptin, which is given to women with breast cancer only if their tumours have a particular genetic characteristic that increases the probability that the drug will be effective. But there are few other examples of pharmacogenetics in actual medical practice. Little is known about which genetic variations predict response to which drugs, and tests are not widely available. Even when such information is known, drug companies have been somewhat reluctant to pursue the practice, possibly because it could potentially confine their drug to a subset of the population, reducing potential sales.
Richard Weinshilboum, a pioneer in pharmacogenetics at the Mayo Clinic, says previous consultations with pharmaceutical companies have provoked a very strong resistance to the development of a test to be marketed with their drug. Recent safety problems with some widely used drugs may have created a shift in attitude and it is becoming clear that some cancer drugs in particular seem to be more effective when directed at tumours with particular characteristics. Steven Paul, executive vice president for science and technology at Eli Lilly & Company, speaking at a biotechnology industry meeting in San Diego, said his company was very excited by the FDA's guidelines.The new guidelines are final versions of a draft issued in November 2003 and apparently differ little from that draft, in the main, the final guidelines eliminate uncertainty.
They do not deal with the use of such tests by doctors in the field but rather the drug companies' submission of data to the FDA Submission of data would be voluntary and the information would not be used in regulatory decisions.
The exceptions might be when a drug company confines the clinical trial to patients with a particular genetic marker, that information then would have to be part of the application for the drug's approval. Submission of information on so-called valid biomarkers, genetic variations whose correlation with drug response is well understood and widely accepted, will be mandatory.
A good example of such biomarkers is the variations in a gene called CYP2D6 that correlate with how quickly people eliminate some drugs from their body. Those who metabolize quickly might need a higher dose and those who do so slowly might get by on a lower dose. The FDA recently approved a test from Roche that measures variations in that gene and in another one involved in drug metabolism. Richard S. Judson, chief scientific officer of Genaissance Pharmaceuticals, a company in New Haven that develops drugs and pharmacogenomic tests, says companies will feel more comfortable with these new guidelines.