Iceland is famous for its volcanic deserts, smoking hot springs, glaciers, and ancient sagasbut not for its high-tech research. Yet on February 2, 1998, the Swiss drug manufacturer Roche Holding of Basel announced it would pay $200 million over the next five years for research by an obscure firm in Reykjavík called deCode Genetics, which at the time had no products, no clients, and less than a year's operating experience.
What does Roche hope to get from deCode? The answer can be found in the faces of Iceland's mostly blond, blue-eyed, extremely homogeneous population. Roche wants Iceland's genes. In the isolated, sparsely settled country, biomedical researchers have found a rich, pure genetic lode, relatively untainted by outside influence for hundreds of years. DeCode is offering itself as the genes' broker and interpreter.
This partnership is just one of many attempts to profit from a revolution in human genome research. A field once dominated by academics has exploded in the past five years with go-go biotech entrepreneurs and pharmaceutical giants. They all dream the same dream: They realize that it may soon become possible to understand the genetic processes that cause diseases and, as the genes yield up their secrets, to find ways of isolating and treating them. As Roche spokesman Roland Haefeli explains, his company thinks deCode's research will help it "make drugs."
As they pursue their research, companies want to patent the key gene sequences to protect their investments. Some critics charge that the patenting of genes amounts to robbing the public commons and is an immoral attempt to turn life into property. But deCode seems to have inoculated itself from charges of exploitation. Roche's $200 million investment is a huge boost to Iceland's economy and, the country hopes, just the beginning. How can anyone get angry about deCode's patenting of Iceland's genetic history when the whole country appears to have bought into it?
The 270,000 people who live in Iceland, many of whom will contribute their genes to this project, are descended from a small number of original settlers, mostly Norsemen who came to the island around the 9th century. Since then, most Icelanders have been intermarrying and choosing their spouses from the same small group of Nordic families.
In the early 1400s, the Black Death swept through the island, killing two of every three inhabitants. Later, smallpox struck. And in the late 1700s, the volcano Hekla, east of Reykjavík, erupted and spewed ash over gardens and pastures. A severe famine followed. These catastrophes, combined with the isolation of the place, created population "bottlenecks" that constricted an already narrow gene pool.
This same harsh environment, however, also made Icelanders an ideal subject for genetic studies. As deCode's president and CEO, Kari Stefansson, says, "We are, in a sense, mining the consequences of natural disasters." The homogeneity of the population will help researchers identify the genes associated with a particular disease more quickly, since a limited genetic sample reduces the background noise. Ethnically diverse populations, such as those in any U.S. city, are difficult to study because they have so many genetic variations that it's difficult to discern which contribute to disease.
Stefansson, 48, a neurogeneticist and an Icelander himself, formed the idea for deCode while teaching at Harvard Medical School. His plan brought the company $12 million from American and European venture capitalists. He then recruited managers among Icelanders, many of whom had been trained at U.S. universities, and opened the lab in early 1997, announcing some ambitious goals: DeCode planned to explore 25 to 35 common diseases.
The company began by identifying the genetic sequence responsible for a syndrome known as essential tremor (shaky limbs). DeCode later studied multiple sclerosis, and laid plans to cover familiar illnesses such as alcoholism, inflammatory bowel disease, colon cancer, diabetes, heart disease, and schizophrenia.
Patients with these illnesses interact with deCode through a network of collaborating physicians. Based in offices around the island, the doctors gather blood samples and provide the company with raw biological material. Before sending in the blood for testing, the doctors remove the patients' names, replacing them with encrypted IDs. After receiving the samples, deCode processes the DNA in its laboratories, obtaining genotypes (genetic profiles) for each individual.
These are then matched with medical records, linking genotype to phenotype dataphysical details, including a person's disease status, age, and weight. DeCode fits each genotype into a jigsaw puzzle of family inheritance patterns that can be verified, if necessary, by referring to Iceland's genealogical records dating back 1,000 years (and now almost entirely computerized). Such a massive cross-referencing should give deCode an unprecedented ability to isolate genes.
The company is also building its own internal database on a large number of Icelanders. DeCode hopes to sell access to this database, called the Genotypes, Genealogy, Phenotypes, and Resources (GGPR) collection, to drug companies such as Roche. The data, deCode claims, will be used only "to identify families in which specific diseases occur, trace the inheritance of the disease over several generations, and rapidly identify the genetic basis of the disease."
Any commercial drug or gene-based diagnostic test developed from the research, Stefansson says, will be provided free of charge to all Icelanders during the lifetime of the patent (between 17 and 20 years). These gestures acknowledge what Stefansson refers to as the company's "core asset." He notes: "[Our] relationship with the population is the most important thing we have."
The country's politicians and medical leaders have given deCode their approval. So has the Health Ministry's Medical Ethics Committee. Iceland's prime minister, David Oddsson, considers the project "extremely important," saying it will help his nation "secure foreign investment."
Stefansson also has taken steps to allay concerns about invasion of privacy. He promises patients who consent to this research that their identities will not be transmitted to the company at all. While deCode retains a master index that links names and encrypted IDs, Stefansson says it is "kept in a safe in our company in an extremely well-shielded room. You need two keys to get into it. We have one of them, and the Icelandic Data Protection Committee has the other." DeCode could never get access to the names without an official watching over its shoulder.
This pledge, and Icelanders' traditional trust in their leaders, seems to have banished skepticism. It's hard to imagine any U.S. agency being entrusted to keep so much volatile information in one databasemedical records, genetic test results, and family histories going back a millennium. But in Iceland, that is exactly what deCode is doing.
While gene collection is moving along at a rapid pace in Europe and the United States, some attempts to collect genes from less developed countries have backfired. Genetic researchers, sponsored initially by the National Institutes of Health and the Department of Energy, have been building up vast archives of human genes for basic research on diseases. Their attempts to conduct a survey of global human diversity came to a standstill, however, when they ran into opposition in the developing world. In China, citizens objected when Western companies made efforts to collect genes using local physicians as research partners, accusing the firms of "biopiracy"quietly exporting valuable genetic data and not sharing the wealth with the local population. In 1997, reacting to fears that China's genetic heritage might be ripped off, the government drafted new regulations limiting the export of biological materials.
But interest from private industry has only accelerated (see chart). For example, in 1993 scientists with Sequana Therapeutics descended upon the tiny Atlantic island of Tristan da Cunha, where the 300 or so residents have an abnormally high rate of asthma. Since then, Sequana (now Axys Pharmaceuticals) has filed patents for what it believes are isolated asthma genes. While Axys could reap millions for a drug that cures asthma, the people of Tristan da Cunha won't have access to it unless they pay for it.
DeCode's project in Iceland is fueling a broader concept than just identifying disease genes. "Pharmacogenomics" is a radically new idea for designing drugs. Adopted last year by many of the largest drug companies, it envisions using genetic testing to screen out the most suitable patients for clinical drug trials. Researchers would select the candidates whose genetic profiles showed that they belonged to a group of individuals most likely to benefit from the drug being tested. If the experimental medicine proved successful, the company would then sell it to patients in the general population with the same genetic characteristics as those in the drug trial. It sounds good: Pills and people could be matched efficiently, reducing bad side effects.
Roche, one of the world's top manufacturers of medical diagnostics, and other pharmaceutical companies have another reason for being interested in matching patients to drugs: It might save a lot of money in clinical trials. If it were possible to screen out the nonresponders in advance, patient enrollment in a typical drug trial might shrink significantly. Since companies pay out of past profits for such trials, this savings would be worth billions of dollars a year.
At least one biotech executiveWilliam Haseltine, CEO of Human Genome Sciences in Rockville, Marylandhas said that he doesn't think it is wise to use genetic screens in order to slot patients into different types of drug therapy, worrying that patients might be screened improperly, and thus be given the wrong, possibly dangerous, drugs.
Another, perhaps bigger, obstacle for this new industry is how patients will react to the idea of genetic testing. If drugs are developed, say, to screen people thought to be at risk for early heart disease, these still-healthy people will have to be genotyped. People will, essentially, need to have bar codes. But they may not want to submit to the testing. They may prefer not to knowor have others knowexactly what risk categories they fit into. Many fear that a centralization of such knowledge could lead to genetic discrimination. If it fell into the hands of personnel offices and insurance companies, for example, they might try to use it to screen out high-risk applicants.
At the moment, there's no immediate problemthe schemes for genotyping patients and tailoring drugs are still only theoretical. But with so much pharmaceutical company money pouring into the hands of researchers, the science won't just be abandoned. Whole-population genetic screening, and the testing of drugs matched to "appropriate" patients, is the brave new frontier in genetic medicine, and it seems likely to get started soon in Iceland, where the population's faith in its leadersand deCodewill ultimately be put to the test.
Eliot Marshall is a staff writer for Science magazine.