June 18, 2002
By NICHOLAS WADE
REYKJAVIK, Iceland — When Kari Stefansson was growing up, Iceland could afford so few imports that fruit appeared in the stores only once a year.
"I remember the smell of apples in the stores at Christmas," he says. "For a thousand years we were desperately poor. And in spite of that we were convinced we had a great culture. This nation lived by that, by the sagas. That may be one reason genealogy became so important."
By way of the University of Chicago and Harvard, Dr. Stefansson, has now returned to Reykjavik, seeking to discover beneath Icelanders' family tree the roots of human disease, and to build a pharmaceutical company from the country's genetic endowment.
The grand design flowed from his professional interest in brain diseases. As a neuropathologist at Harvard, he reckons that he "cut 10,000 brains." A disease of particular interest to him was multiple sclerosis. His native Iceland seemed an ideal place to explore its possible genetic roots. Besides a wealth of genealogical records, Iceland has comprehensive medical records and cooperative patients.
But when he applied to the National Institutes of Health for a grant to study multiple sclerosis in Icelandic families, he was turned down. At that time geneticists were convinced that only sibling studies — patients compared with their unaffected brothers or sisters — had the statistical power to uncover the roots of complex diseases.
Distantly related members of Icelandic families would not suffice to show up a link between genes and disease, he was told.
"If it weren't for N.I.H. rejecting our grants there would be no Decode today," said Dr. Jeffrey Gulcher, a Harvard colleague who joined the company as its research director.
Dr. Stefansson decided that his project would have to be carried out with private capital. Within six weeks he had raised $12 million, founding Decode in 1996.
He prevailed over the next obstacle, strong protests from geneticists in Iceland and abroad alleging that Icelanders were being exploited and their privacy put at risk. After an open, somewhat rancorous debate, the idea was embraced.
Now, the company says, 90 percent of the patients it invites to join a study accept, and of these 99 percent re-enlist if asked to be studied for a second disease. Many seek to contact Decode to learn more about their genetic condition, but privacy requirements bar any such contact. Privacy advocates, Dr. Stefansson says, have decided on these patients' behalf that their privacy is more important to them than their medical welfare.
Creating Decode has required making a lot of hard calls correctly. "I have a lot of very talented people and well-educated people around me who voted me down when I wanted to make the wrong decisions," Dr. Stefansson said.
But he has been the central figure in making Iceland the focus of a hunt for disease genes, in persuading Icelanders to accept the idea and in building new departments on the forefront of new disciplines like statistical genetics and bioinformatics.
A striking figure often dressed in black that offsets his shock of white hair, Dr. Stefansson, 53, presides over his growing creation with a distinctive brand of pleasantly sardonic humor. Decode is still small enough for him to know the family histories of many employees, review every manuscript and, one day last month, take a succession of phone calls from a teenage daughter in need of genetics tutorials for an imminent exam.
Dr. Stefansson sees population genetics and computers as part of a new world that medicine has yet to enter. "My prediction is that in 10 to 15 years every society will have these databases," he says.
In one recent vindication of his ideas, the Wellcome Trust of London said in April that it would start the UK BioBank, a plan to analyze the genetic roots of disease in the English population.
Knowledge of a population's genetic risk of disease is just one step. Another is computerized health care databases of the kind he has started in Iceland.
"Every car mechanic today has software to help in diagnosis and decision making, though a car is infinitely simpler than a human being," Dr. Stefansson said, then added in disbelief, "Physicians still write medical records with a pen."
The young company he has built stands within reach of success. Reports on genes whose variations cause three complex diseases are under review at scientific journals. Several pharmaceutical companies have signed up for a special service Decode offers, that of testing genetic reaction to drugs, a trait that can be mapped just like disease genes. Applied Biosystems, a leading maker of DNA sequencing machines, has adopted Decode's genotyping software.
But Dr. Stefansson still sees resistance to overcome. Like the 10th century poet Egil Skallagrimsson from whom, like many Icelanders, he claims descent, he always has another struggle ahead. He believes the scientific reports he is about to publish will show the Decode experiment has succeeded in ways that others have not.
Through its detailed knowledge of Icelandic genomes, for example, Decode has discovered and been able to correct many large-scale errors in the human genome sequence produced by the public consortium of academic centers, Dr. Stefansson and colleagues report in the current issue of Nature Genetics. In more than 100 cases, large sections of the consortium's human genome are in the wrong order or flipped head-to-tail, according to Decode's new genome map, and the wrong order degrades the statistical power of gene-hunting methods.
"Here we are, a tiny little biotech company on a rock in the North Atlantic, and we have put together a map than can vastly improve the human genome," Dr. Stefansson said, noting that the architects of the Human Genome Project had not ordered up such a map themselves.
"These guys came out of the discipline of genetics and went into this massive nucleic acid project and they forgot their origins," he said, "not realizing genetics could have helped them dramatically in the assembly of the genome sequence."
A spokesman for the National Institutes of Health's genome office said the project had made use of two genetic maps that were available.
Decode's pending report on a variant gene that causes schizophrenia will be a notable achievement, if valid. A Decode team under Dr. Hreinn Stefansson (no relation to Kari) found a strong linkage between schizophrenia and a region of Chromosome 8. Several teams working in other countries have identified the same general region but so far have been unable to pinpoint which of the many genes in the region is responsible. The Decode team believes it has found the gene, which is called Neuregulin 1 and, though quite well known, has not before been linked to schizophrenia. Its role seems to be to control the number of synapses, or connections between brain cells.
If Decode shows it can indeed pinpoint the variant genes behind common
human diseases, Dr. Stefansson will have succeeded on many levels. He will
have developed a powerful general method for linking the human genome sequence
to disease. He will have started a new industry in Iceland. He will have
put Icelanders first in line to benefit from the discoveries made with
their genetic inheritance. And a cold wet rock long at the edge of history
will suddenly loom larger on many maps.
Copyright 2002 The New York Times Company