Snipping away at human disease
Into the era of individually tailored therapies

By Joannie Fischer

Warren Wegele would probably be dead by now if he had received the standard treatment for his failing heart. Wegele is one of the many patients who expose the shortfalls of medicine in all areas–from cancer to depression–by not responding to medicine or by having toxic, even fatal, reactions. "Each time I write a prescription for a new patient," says diabetes expert David Altshuler, "I know it might not help at all and could do harm." In fact, properly prescribed medications make 2 million Americans seriously sick and kill 100,000 each year.

But the era of "one size fits all" medication is ending, as physicians learn to read a patient's unique genetic code and to tailor treatments accordingly. In Wegele's case, doctors at the University of Cincinnati found an alteration in one of his genes, alerting them that current drugs would do no good; he received a new heart instead. A massive effort now underway to catalog the variations in DNA is ushering all fields of medicine into what researchers call the "post-genomic era," when DNA tests will be as routine as blood pressure and temperature checks.

Now that the Human Genome Project has spelled out the 3.1 billion chemical letters of the average human's DNA code, the hunt is on for what researchers call SNPs (pronounced "snips"), the spots along the chain where spelling differs from one person to the next. Humans are remarkably similar: 999 out of every 1,000 letters are identical among all people. Yet the genetic code is so long that those one-in-a-thousand variations add up to make some people tall, others curly haired, some prone to heart attack, and still others resistant to migraine treatments.

In all, there are probably 10 million SNPs (shorthand for single nucleotide polymorphisms), and decoding how they affect individuals could revolutionize both the diagnosis and treatment of today's most common diseases. Last month, researchers at the SNP Consortium–a nonprofit collaboration of top academic centers and 13 of the world's largest pharmaceutical companies–announced that they have found more than a million SNPs so far. Working independently, other pharmaceutical firms also boast millions of SNPs in their private databases. Altshuler, who is part of the consortium and helped to develop a speedy way of finding SNPs, says that enough SNPs have already been identified to help researchers better understand most of today's common killers.

Some physicians are trying out this new knowledge in the clinic and meeting with early success. For example, at the University of Cincinnati, physician Stephen Liggett and colleagues studied the SNPs of 121 asthma patients who came in for treatment this year. Most often, the drug Albuterol is used for asthma. It affects a particular gene–the so-called (beta) 2 adrenergic receptor–which in turn affects lung function. There are 13 different known SNP areas on this gene, which could theoretically be combined to form 8,192 different SNP patterns (called "haplotypes").

Fresh air. While no single SNP could affect a person's reaction to Albuterol, Liggett found that four of the haplotypes did affect response. Depending on which of those four SNP groupings patients carried, doctors could predict whether or not they would respond to the drug. In this study, one third of patients needed an alternative treatment. The study is preliminary but promising for the 17 million people with asthma and for drug makers eager to develop new products that will target those who don't respond to current treatments.

But before the new field of "pharmacogenomics"–designing drugs to match particular DNA profiles–can truly take off, researchers need to better understand how the many SNPs on the many genes involved in any given disease all interact. There are hundreds of genes involved in diabetes, for example. And given the complexity of even a small stretch of DNA, other common diseases may prove even more difficult to unravel. Last month, scientists at the Sanger Center in Cambridge, England, reported in the journal Nature that they had found 2,730 SNPs on chromosome 22 (one of the body's smallest chromosomes), which has been linked to 35 disorders, including schizophrenia and heart disease.

With so many factors involved, researchers will have to devote serious effort to sorting out the truly significant SNP groupings from mere "noise." When researchers at the Whitehead Institute Center for Genome Research at the Massachusetts Institute of Technology recently followed up on 16 SNPs reportedly linked to diabetes, rigorous studies of thousands of samples were only able to confirm that one of those 16 is truly linked to the illness.

A better map. Most researchers agree, though, that it is just a matter of time until the connections between SNPs and disease become clearer and the true causes of many modern ailments are uncovered. "Right now, we just treat symptoms," says Eric Lander, director of the Whitehead Institute. But the SNP map, Lander says, will explain the genetic underpinnings of illness. In fact, our very definitions of some disorders are likely to change, he says. What is now thought of as one disease, such as multiple sclerosis or breast cancer, could be shown to be three or four distinct ailments, each needing a different treatment. Ultimately, it may be possible to actually correct the genetic spelling errors that underlie cancer or Alzheimer's through a process known as gene repair.

But the era of personalized medicine cannot materialize until more people are willing to have their DNA scanned for links between genetic variation and disease. Now, "no one wants to participate" in DNA studies for fear of losing his privacy, says Arthur Holden, chairman of the SNP Consortium. Last week, Holden announced that he is forming a partnership with IBM to launch First Genetic Trust, a gene bank that will guarantee the privacy of anyone who donates a DNA sample. If authorized, the bank will send the sample to an individual's physician or to a study that could benefit.

Because SNPs are involved not only in illness but in every human trait, many worry that knowledge of SNP variations could be used to discriminate against people or whole groups of people. But Altshuler says that the more DNA he studies, the more amazed he is by how much everyone has in common. "If you take any two people from 100 in an auditorium, they will vary from each other," he says. "But if you then take the other 98 people, they are likely to have the same variations as one of the first two. We quite literally are one big family."