The Age of Personal Medicine Begins

Posted on March 18, 2010

By Dr. Jim Logan

One of the predicted benefits from the Human Genome Project, the 13-year international research effort to sequence human DNA, was the promise of ‘personal medicine.’ Though slow to materialize, the dawn of personal medicine has finally arrived. This is good news for anyone who ever has taken a medication - - or ever will. That’s just about all of us.

The sequence of base pairs in human DNA determine the order of amino acids incorporated into the proteins that comprise the metabolic machinery of our bodies. Special proteins called enzymes catalyze the chemical reactions that convert the medications we take into a series of metabolites. Some of these metabolites are the biologically active forms of the drug, some cause dangerous side effects and some interfere with other drugs. The metabolic pathway is more of an interlocking web of reactions rather than a simple linear sequence of reactions. Each step in the pathway is facilitated by a different enzyme. The ultimate goal of drug metabolism is to convert the original medication to a form that can be cleared from the bloodstream and excreted.

Physicians and scientists have known since the 1930s that different people can metabolize the same drug differently. Doctors assumed, correctly, these differences have something to do with the presence and severity of side effects. We now know these differences are genetically determined. There are multiple forms, so-called alleles, of the same enzyme. Even though they catalyze the same reaction, each allele exhibits a slightly different enthusiasm or affinity for the task at hand. Some are so unenthusiastic, they don’t work at all and are considered deficient or defective.

Drug side effects, termed adverse drug reactions (ADR), are not trivial. A large meta-analysis of hospital patients in 1998 found ADRs are responsible for a whopping 106,000 deaths, 2.2 million “serious medical events” in the US each year and cost about as much as the drug treatments themselves.

If you compare the list of drugs most commonly known to produce ADRs to the list of those enzymes known to exist in multiple forms, so-called polymorphisms, the overlap is stunning. The drugs most commonly implicated in ADRs are also the most likely to be metabolized by enzymes known to exhibit polymorphisms.

To some degree physicians have always used personal information (age, race, gender, family history, etc.) in an epidemiological sense to crudely predict a patient’s response to therapy. But now it looks like we may be able to take personal medicine to a whole new level. If, by sequencing an individual’s DNA, we can determine the specific alleles a patient has inherited, is it possible to predict (and therefore potentially prevent) the incidence and severity of ADRs in that patient?

Turns out the answer is a qualified YES. Enter pharmacogenetics, the study or clinical testing of genetic variation that gives rise to differing response to drugs, and pharmacogenomics, the more general application of new genomic technologies to find new drugs and better characterize older ones.

In 2004, 6-Mercaptopurine (6-MP), a drug used to treat a form of leukemia (the most common type of cancer in children), became the first drug ever to include special instructions in the label to permit adjusted doses for each patient. Thankfully these drugs are very effective with reported cure rates of up to 85 percent. But sometimes they can be associated with very dangerous side effects. For over a generation scientists couldn’t figure out why some children had adverse effects while others didn’t. The answer was, of course, genetics.

The enzyme that metabolizes 6-MP is thiopurine methyltransferase (TPMT for short). TPMT has multiple alleles. One allele results in a deficient enzyme which means 6-MP metabolism must proceed along another pathway. This alternate pathway unfortunately produces a metabolite that is very toxic to the bone marrow at high doses. Those patients who have two deficient alleles (one in three hundred people) exhibit an almost complete TPMT deficiency. They only need 6-10% of the standard dose of 6-MP. If they are treated with the full dose, they risk severe bone marrow suppression, a life threatening condition that can (and does) result in death. The bottom line: their genotype predicts clinical outcome, a prerequisite of any effective pharmacogenetic test.

Unfortunately it’s not always this straightforward in medicine. I gave a “qualified YES” above because in the case of 6-MP, those individuals who are heterozygous (they carry one of three ‘variant’ alleles and one normal allele) produce a reduced quantity of TPMT, the functional enzyme. Even though they are at greater risk of adverse effects, their genotype is not necessarily predictive of clinical outcome.

6-MP may have been the first but there are other even more commonly used medications for which the specific interactions of polymorphism and adverse drug reactions have been determined:  Tamoxifen (a drug used to protect against recurrence of breast cancer); Plavix (the block-buster anti-clotting drug); and Warfarin (the widely prescribed blood thinner). It is very likely drug labels for these popular medications will be modified soon to permit individual dosing.

In the next five years, the cost of sequencing a person’s genome will likely drop below $1000 according to Dr. Francis Collins, chief scientist for the Human Genome Project. That’s about the cost of a colonoscopy. When it does, pharmacogenetics will become mainstream.

One prediction I can confidently make: I will be the first one in line. Perhaps I’ll see you there.

Comments