The Making of the Pharmacogenomic Prescription

The appearance in 1998 of a new medicine for an intractable and especially savage form of breast cancer was a medical milestone for two reasons.

For one, the drug, called Herceptin, shrank tumors and prolonged lives.

What received much less attention, however, was the unique way in which Herceptin is prescribed.

Herceptin is one of the first drugs for which tests are performed to predict whether it will work in a particular patient prior to drug prescription.

How do physicians know if Herceptin will work in a particular patient?

The drug is specially designed to treat metastatic breast cancer patients whose tumors are shown to express abnormally high amounts of a protein called HER2.

For those patients – up to 30 percent of women with breast cancer – Herceptin can bind to HER2, slowing tumor growth.

The flip side: for those with normal HER2 levels, the drug is as useless a weapon as a hilt without its blade.

That Herceptin and other drugs help some individuals but not others is not exceptional – in fact, it is the rule for drugs. This principle was established around 40 years ago by researchers studying pharmacogenetics, the science of how genes influence response to medicine.

Pharmacogenetics revealed instances of genes influencing drug responses in enormous numbers of people.

Millions, for instance, get less pain relief from codeine because their particular variant of a gene called CYP2D6 is unable to convert codeine into its active form, morphine.

The codeine/CYP2D6 story is not alone.

If they take aspirin or one of dozens of other drugs, more than 400 million inhabitants of equatorial Africa, Asia and South America are at risk for hemolytic anemia, a potentially fatal reduction in the blood’s capacity to carry oxygen, due to differences in another gene, called G6PD.

Despite these cases, insights were few and difficulties of routine medical testing for gene variants were great. So, pharmacogenetics remained for quite some time a minor field of research.

Now the human genome project, the international effort to catalog all human genes, has brought the idea of personalized medicine into the limelight.

Renamed “pharmacogenomics” to reflect a more global genome survey and revamped to supply actual medical applications, the field has a simple goal: to develop genetic tests that will help doctors prescribe drugs that work, not ones that don’t.

Major drug companies, joined by a young band of pharmacogenomics companies, are hunting assiduously for variants of human genes that explain why drugs work well for some but not others.

Consider how better-suited prescriptions might improve treatment of high blood pressure, says Bill Campbell, a specialist in clinical laboratory tests and Director of Genomics at Princeton, NJ-based Covance, a drug development services company.

“Today,” Campbell explains, “if a patient has high blood pressure, a doctor will say, ‘This is a pretty good drug. Why don’t you try it at this dose.’ When the patient is tested a few weeks later, if the drug doesn’t work, the doctor may try increasing the dose. Then the patient has to return later to see if that works. If it still doesn’t, the doctor will have to try something else. It may take six or seven visits before the right combination of drugs is found.”

The point of pharmacogenomics is to avoid all that by getting the prescription right the first time.

Trial and error exacts an even higher price with immediately life-threatening diseases like certain cancers, where using the wrong drugs means patients may not survive to try something new.

“You can’t afford to let someone suffer six months of irreversible decline because they didn’t get the right therapy,” says Colin Dykes, chief scientist of Variagenics, a Cambridge, Massachusetts based pharmacogenomics company.

But delay and suffering is precisely our situation today when, according to Taylor Crouch, the company’s CEO, most cancer drugs help less than half of patients, exposing them to toxic side effects without hope of improving their health.

Herceptin leads a parade of at least four drugs in various developmental stages that attack tumors with overactive cancer genes. Intended for a variety of cancers – brain, melanoma, prostate, breast, and lung- they will be prescribed only after first verifying cancer gene overactivity.

As knowledge increases about how tumor-causing genes influence drug responses, pharmacogenomics may also help doctors use drugs to prevent cancer.

Limited clinical studies show that Tamoxifen reduces risk of breast cancer in women with BRCA1 and BRCA2 gene variants that heighten risk of the disease. This has spurred scientists to ask if similar measures might lower risk of bladder and prostate cancers.

Cancers, of course, won’t be the only diseases affected by pharmacogenomics. Discoveries of gene variants affecting how Tacrin works with Alzheimer’s patients, how Pravochol lowers cholesterol, and how Albuterol helps asthma point to just a few of the diseases in which pharmacogenomics will have a big impact.

Personalized medicines have also been introduced into the treatment of HIV.

In treating AIDS, doctors prescribe combinations of three or four drugs at a time to hold the HIV virus at bay. Although there are now 17 drugs from which to choose combinations, there are also at least 120 HIV gene variants that render one or more drugs ineffective.

To learn which medicines it would be useless to prescribe, doctors have increasingly been getting their patients’ viruses analyzed for drug-resistance gene variants. Ineffective treatment with the wrong drug – and the associated cost and suffering – can be thus avoided.

About the Author

has written 7 stories on this site.

Write a Comment

Gravatars are small images that can show your personality. You can get your gravatar for free today!

Copyright © 2017 Gene Letter.