The Iceland Experiment

By MICHAEL D. LEMONICK

Dr. Kari Stefansson can trace his ancestry back 1,100 years. That's almost unheard of in the U.S., but in his native Iceland, where genealogy is a national obsession, it hardly raises an eyebrow. The island nation is a genetic anomaly: settled by a few Norsemen and Celts in the 9th century A.D. and relatively free of later immigration, it is among the most genetically homogeneous countries on earth. And in the late 1990s, when scientists were racing to map the human genome, Stefansson realized that Iceland's genetic isolation and unrivaled genealogical records made it a potential gold mine for isolating genes.

Thus began Iceland's great genetic experiment, an attempt to mine the gene pool of an entire country in search of the root causes of--and potential cures for--some of the world's worst diseases. And after years of controversy, dashed hopes and burst stock bubbles, the effort is finally paying off. Over the past decade, deCODE Genetics, the company Stefansson co-founded in his home city of Reykjavik, has discovered more than a dozen genes linked to diseases ranging from stroke to schizophrenia. Last month, deCODE announced that it had found a gene that boosts the risk of Type 2 diabetes. And within a few weeks, the company will start the final phase of trials for a drug based on a newly identified heart-attack gene that appears to be especially dangerous in African Americans. "I'm very enthusiastic," says Dr. Francis Collins of the U.S. National Institutes of Health and leader of the Human Genome Project. "What deCODE is doing is clearly exciting, and I congratulate them."

In principle, their method is straightforward: to find a disease-related gene, find someone with the disease, then see how his or her DNA differs from the DNA of healthy people. In practice, however, individual genes rarely cause illness on their own; instead, they tend to make people more susceptible. And in places with genetically mixed populations, the complex interaction among genes makes it hard to find the risky ones. But in Iceland, with its uniform population and genealogies that show how everyone is related, risky genes tend to stand out. The country's meticulous medical records provide even more data.

Ingenious as it was, Stefansson's plan quickly ran into problems. In order to build a database of genomes, deCODE needed blood samples from as many Icelanders as possible, as well as access to their health records. Parliament granted permission to tap into those records, along with an exclusive license to assemble, maintain and market the resulting data. Thousands of citizens donated blood, and many bought shares in deCODE as well. But those shares, which rose to a high of $65 in a frenzied run-up in 1999 and 2000, plunged to as low as $2 in the collapse of the dotcom bubble. They're around $9 today--and deCODE still hasn't turned a profit. Investors lost a lot of money, and the firm was forced to lay off scores of employees.

Then in 1998 the U.S. firm Hoffmann--La Roche agreed to pay $200 million for the right to develop drugs based on some of deCODE's data. The idea that a foreign company might profit from their personal information made many Icelanders balk. A woman named Ragnhildur Gudmundsdottir sued to keep her deceased father's medical records from going into the deCODE-run database, citing a right to privacy, and in 2003 Iceland's supreme court ruled in her favor.

Having lost its guaranteed access to every citizen's records, deCODE had to change tactics and approach people one by one. In return, the company promised that Icelanders will get any drug Hoffmann--La Roche develops out of the project for free until the patents run out. According to Stefansson, most have agreed to cooperate. "Ten percent of people have questions about the project," says Asmundur Johannsson, a Reykjavik resident. "Ninety percent approve of deCODE, and I am one of them."

Thanks to people like Johannsson, a huge freezer in the basement of deCODE's gleaming, modern Reykjavik headquarters now holds blood samples from about 100,000 individuals, roughly half of Iceland's adult population. Using those samples, scientists at the company were able to zero in on their new anti-heart-attack compound. It's based on a gene known as LTA4H, first seen in mice, which governs the production of an enzyme called leukotriene A4 hydrolase. The enzyme plays a role in inflammation, a key factor in heart disease, and also encourages the buildup of cholesterol on blood-vessel walls.

And sure enough, Icelanders with a particular variant of the LTA4H gene turn out to be 40% more likely than average to have heart attacks. Looking outside the country, deCODE scientists found the variant gene in other populations--and discovered that in African Americans the increased risk is not 40% but a whopping 250%. That suggests the company's prospective drug--invented by Bayer and licensed by deCODE--could have a correspondingly large lifesaving effect, although even if it works, it could be several years before it reaches the U.S. market. Some critics are worried that insurers and employers might avoid anyone bearing the bad gene, making discrimination even worse than it already is. Stefansson scoffs at that notion: "You guys never needed genetics to discriminate against African Americans," he says. "You've done that completely unassisted by genetic discoveries."

The idea of combing through populations for disease genes isn't unique to deCODE. Britain's UK Biobank, for example, will follow 500,000 volunteers for decades, trying to correlate genes, lifestyle and disease. And two initiatives being put together by the U.S. National Institutes of Health will look for nearly 20 diseases in up to 40,000 people. But with its long head start and Iceland's genetic advantages, deCODE could be hard to catch. So far the company has isolated 15 gene variants for 12 diseases, including stroke, schizophrenia, osteoarthritis and, most recently, diabetes. In addition to the heart-attack drug, it has medications in the pipeline for preventing asthma and atherosclerosis. Even when no drug is available, knowing you have a disease gene can be invaluable. "What it tells you," says Stefansson, "is whether you are at risk, and it gives you the opportunity to respond. This is liberating."

With reporting by Reported by Helen Gibson / Reykjavik, Alice Park / New York