 
|
|||||||||||||||||||||||||
              
|
Pharmacogenomics: One day, it may allow doctors to tailor drugs to individuals
Sunday, July 09, 2000 By Byron Spice, Science Editor, Post-Gazette
In the future, before reaching for a prescription pad, your doctor may want to consult your genetic profile. Perhaps you need an antidepressant medication, or a drug to treat your allergies, or you are about to undergo cancer chemotherapy. In all these cases, your genetic makeup may have no direct bearing on the diagnosis your doctor makes, but everything to do with which drug he decides to prescribe.
Maybe your liver doesn't make an enzyme necessary to activate a particular antidepressant. Or maybe you're that one-in-100 person who will develop lethal side effects to a chemotherapy drug.
Doctors, pharmacists and most patients already know that not everyone responds to every drug. But no one has been able to do much about it, except adjust dosages or switch drugs if side effects appear or if the patient's condition doesn't improve.
New genetic tools may change all that.
In an approach known as pharmacogenomics, researchers hope to match patients to drugs based on the patients' genetic profiles, and to use genetic information about diseases to guide development of new drugs.
"It is inevitable that the era of individualized medicine is going to come," said Leaf Huang, director of the Pharmacogenetics Center at the University of Pittsburgh School of Medicine. Even the best drugs on the market are only effective in about 80 percent of patients, he noted. Genetic profiling could help identify those patients who will suffer side effects or simply won't respond to a drug.
It likely will be another 20 years before many genome-based drugs hit the market, however. It will take years to analyze the genome, find new targets for drugs and develop those drugs, said Garth Ehrlich, director of the Center for Genomic Sciences at Allegheny General Hospital. "Even once you have a new drug, it takes seven or eight years to get through the regulatory process," he added.
Managed care companies may well lead the push for pharmacogenomics, said Dr. Ronald Bachman, chief of genetics at HMO giant Kaiser Permanente's Oakland, Calif., hospital.
For instance, of all the cancer patients who receive the chemotherapy drug azathioprine, 2 percent will die and 10 percent will be hospitalized because of toxic side effects. A genetic test that would identify the most vulnerable patients would save lives. "And look at the money you're saving on that 10 percent of people" who would otherwise require hospitalization, he added.
Your enzyme profile
A major reason why people respond differently to drugs is differences in metabolism, which often involves a set of proteins produced by the liver known as cytochrome P450 enzymes, said Dr. Bruce Pollock, director of geriatric psychopharmacology at Western Psychiatric Institute and Clinic.
Enzymes are proteins that act as chemical catalysts, speeding up chemical reactions the body needs to function. These particular enzymes are a product of the warfare between plants and animals. Most drugs today are derived from plants, which protect themselves by making various compounds that are toxic if ingested by animals.
The P450 enzymes in people make chemical changes that turn the digested plant products into poisons.
By controlling dosages, doctors have learned to use these poisons therapeutically. But the therapy fails if the patients lack certain versions of the P450 enzymes.
About 10 percent of people lack the 2D6 version. Without it, codeine is no better than a placebo. Many Asians lack the 2C19 enzyme needed to metabolize Valium. A shortage of the 3A4 enzyme proved dangerous to people who took the antihistamine known as Seldane; the drug was withdrawn from the market in 1997 when interactions with drugs known to inhibit the enzyme resulted in patient deaths.
In a study set for publication in a scientific journal, Pollock and his colleagues have shown that genetic differences can affect how people respond to the antidepressant Paxil.
Like Zoloft and Prozac, Paxil is a selective serotonin reuptake inhibitor, or SSRI. People with too little serotonin, which is a brain chemical, have a tendency to be depressed. SSRIs increase levels of serotonin by preventing nerve cells from reabsorbing it after it has been released. But Pollock found that some people are genetically inclined to have extra substances called serotonin transporters; their brain cells thus are better able to reabsorb serotonin, reducing Paxil's effectiveness.
If doctors knew which people had the genetic profile for extra serotonin transporters, they could raise the dosage of antidepressant medications to compensate, Pollock said.
Targeting cancers
Another area where genetic profiling might make a big difference is chemotherapy.
Pittsburgh-area neurosurgeons, for instance, have formed a research consortium to find a more rational way of targeting brain tumors. Dr. Robert Selker, chief of neurosurgery at the Western Pennsylvania Hospital, said the choice of chemotherapy is largely hit-or-miss for brain cancers and the results, not surprisingly, have been poor.
Now, he and his colleagues are trying to test tumor samples in the lab against various chemotherapy agents. They also are looking for genetic clues, such as the presence of the her2 receptor on cell surfaces that is targeted by the cancer drug Herceptin, to help them select chemotherapy drugs for each patient.
Dr. Arthur Feldman, Pitt's chairman of cardiology, said his department routinely performs genetic analyses of heart failure patients. It hasn't had any effect on their care yet, but he hopes to develop cocktails of heart failure drugs that can be offered patients with specific genetic profiles. Many heart failure drugs, he noted, depend on P450 metabolism.
But as much as physicians would like to use genetic information to optimize their use of existing drugs, "the pharmaceutical companies don't want to do these studies," Feldman said.
If a company has a drug on the market that is approved for all patients, it has little motivation to fund studies that might prove it is effective in only 60 percent of those patients.
On the other hand, Pollock noted, drug companies are very interested in using pharmacogenomics to develop new drugs, and are investing billions of dollars to find new medications that will target specific genes or proteins involved in a disease.
Last year, 10 of the world's largest pharmaceutical firms joined together in the SNP Consortium, a two-year, $45 million effort to identify thousands of cancer markers -- known as single nucleotide polymorphisms -- that might be used to establish genetic profiles of patients.
The SNPs are single-letter spelling variations in the genome, and while they may not directly cause disease, they may be located close to defective genes that do play a role in cancer or some other problem. Researchers hope to use SNPs to distinguish between different subtypes of diseases, or identify patients with different tolerances for a particular medication.
The Human Genome Project, which is mapping the entire genome, also has its own SNP search under way.
A high-cost search
James Pierce, assistant professor of genetics and biotechnology at the University of the Sciences in Philadelphia, said planned pharmaceutical mergers announced in the last year -- Glaxo Wellcome with SmithKline Beecham, Pfizer with Warner-Lambert, Pharmacia & Upjohn with Monsanto -- were driven in large part by their hopes of cashing in on pharmacogenomics.
"The technology is expensive, the talent is expensive," explained Pierce, formerly senior scientist at Magainin Pharmaceuticals Inc. The payoff for genetically targeting drugs will be long-term, he added, but the expenses are short-term and difficult for all but the largest companies to bear.
One payoff, Pierce said, may be in reducing the cost of clinical trials, which now top $300 million for each new drug and are a major contributor to the costs of new drugs. The hope is that genetic profiling may allow researchers to screen out patients who are likely to have a bad reaction to a medication.
"Imagine the benefit to the development of new therapies if drugs entering clinical trials are almost ensured to be well tolerated in the body and to have the desired effect," Chris Sander, chief information science officer for Millennium Pharmaceuticals, wrote in the journal Science this spring. "Or imagine relatively short clinical trials, [as] confirmatory final tests to guarantee that drugs and diagnostics are safe and effective."
One consequence, however, is that not every patient would qualify for every new drug.
"In the future, certain drugs won't be considered safe to use without first doing a genetic test" to screen out patients who would suffer side effects, said Dr. Joseph Scherger, chairman of family medicine at the University of California, Irvine. Drug companies may need to develop, say, five different drugs to treat allergies in different subsets of patients.
But William Haseltine, chairman of Human Genome Sciences, a Rockville, Md., pharmaceutical firm, says that's not likely.
"Tailor-made medicines are absurd to the point of being ridiculous," he said, maintaining that no firm, already reeling from the costs of developing new drugs, would seek to develop five complementary drugs to treat a single disease.
The goal should continue to be one drug for one disease, Haseltine said, a strategy his company is pursuing by analyzing genetic activity in the laboratory to identify and produce natural hormones, proteins and antibodies.
|
||||||||||||||||||||||||
