During the late 1990s, it was an article of faith that the decade ahead would be the Age of Biotech, an era when science would unravel the mysteries of DNA to determine the causes—and potentially the cures s—of the world's most common ailments: heart disease, cancer, mental illness. But even in an industry accustomed to hype, one company stood out: a small Icelandic startup called deCODE. In 1999 The New Yorker ran a glowing 10-page profile with a portrait of Kári Stefánsson, the company's charismatic founder, climbing the DNA double helix like a ladder. DeCODE, it said, had "a scientific instrument of unparalleled power." Over the next 10 years, deCODE discovered so many genes linked to diseases that its only competitors were top U.S. labs lavishly funded by the National Institutes of Health. What it didn't find, however, was a business model that made sense—a common problem in the speculative, science-driven world of biotech. That, plus the global financial crisis, caused it to burn through $676 million, and last November it went bankrupt.
DeCODE's story is dramatic, but these days it's a common narrative arc. The recession has been tough on every company, but it's been particularly virulent toward startups, which usually lack profits and capital, giving them little margin for error when revenue falls and credit tightens. Typically, when a startup goes bust, it's gone forever—but deCODE hopes to be the exception. In the last two months, it's been recapitalized by outside investors, and now the company is emerging from bankruptcy with new leadership and a new strategy. There's still much that can go wrong: its new model isn't proved; there are early signs its new CEO's strategy may conflict with its visionary founder, who remains on the scene; and the company's culture, which has always shown more reverence for science than for profits, will need an overhaul. But in an environment where few companies of its size get a shot at resurrection, deCODE's second act is shaping up to be one of the biotech industry's most interesting experiments.
To say that deCODE launched big is like saying its home country of Iceland gets a bit chilly. As a group, Icelanders are closely related, which means when they get sick, there's a limited subset of genes that can be blamed. They also keep extensive genealogical records. Stefánsson's insight was to put their genes and genealogy together and search the results for the roots of disease. Initially, he hoped to analyze the medical histories of all 277,000 of his countrymen through an exclusive government deal. That plan alarmed some bioethicists, who accused the company of profiteering and putting its subjects' privacy at risk. (Stefánsson's dismissal of their concerns in The New Yorker as "a horrendous crock of s--t" probably didn't help his case.) Ultimately, the plan was ruled unconstitutional. Still, deCODE was able to gather genetic information from 140,000 Icelandic volunteers, and its database made it one of the world's top players in biotech.
What deCODE did with its data was the first step toward its financial downfall, though no one suspected it at the time. The company's original plan had been to focus on making diagnostic tests, but investors were lukewarm tothat idea. Diagnostics, which use genetic material (like a cheek swab or blood sample) to test for the presence of certain genes, are tricky to bring to market, and at the time, the technology for making them was somewhat primitive. The way venture capitalists saw it, drugs were a more proven route to profits. If deCODE could discover even one gene with a large effect on a common condition like heart disease, the thinking went, it would provide an obvious pharmaceutical target and a way to rake in cash. So in 1998, deCODE signed a $200 million drug-discovery deal with F. Hoffman–La Roche. Then it spent the next 11 years linking gene after gene to major diseases. "Kári published so many breakthroughs in [the British science journal] Nature that it might as well have been his personal blog," says Balaji S. Srinivasan, chief technology officer at Counsyl, a prominent genomics startup. "They were the NIH of Iceland."
The trouble was that unlike the government-funded NIH, deCODE needed to turn a profit. Despite years of spectacular scientific success, it never did. The genetics of illness turned out to be more complex than researchers expected. At deCODE and elsewhere, the new genes linked to common diseases turned out to be rare or to have only small effects on individual risk. That killed any prospect of using deCODE's discoveries to make blockbuster drugs. In 2007, hoping for a new revenue stream, the company introduced deCODEme, a $985 personal genetic test that was part medicine and part recreation. But competitors, such as 23andMe, cofounded by Ann Wojcicki, wife of Google's Sergey Brin, quickly crowded that market, and sales fizzled.
Then came the financial crisis. DeCODE had unknowingly made another fatal mistake by handing all its operating funds to Lehman Brothers, which put the cash into speculative bonds that went bust. By November 2008, the company had lost most of its funding. Its stock, once traded at $31.50 a share, hit a low of 19 cents and fell off the NASDAQ. The Icelandic government, hugely supportive of deCODE in the past, could not help out—it, too, was going broke. On Nov. 17, 2009, deCODE filed for Chapter 11 bankruptcy in the U.S. It started trying to sell off the few drugs it had begun to develop and put its Icelandic gene-finding operation on the auction block.
Now, with capital from Saga Investments, deCODE is gearing up for its second act. New CEO Earl "Duke" Collier is trying to impose financial discipline and keep the focus on marketable strategies to ensure deCODE doesn't go belly-up again. His first goal is to turn the company into a sort of consulting firm. New techniques for sequencing enormous amounts of DNA have recently deluged researchers with data; analyzing that data and "giving it back in a nicely packaged way" will be a useful service, he says. But it will be challenging to refocus the company without harming its freewheeling, academic atmosphere. "Discovery calls for a certain creativity and chaos," says Stefánsson, now 60. "We have the most productive discovery operation in the world, and we do not want to change that."
Given that deCODE was Stefánsson's baby, Collier, 62, also has to teach his new corporate partner to co-parent. At times, the two work well together, like a later-life version of Brin and Larry Page. "Kári is something of a savage, and I'm more polite," says Collier, "but underneath we're very much the same." Years as an executive VP at Genzyme and a member of deCODE's board have taught Collier some lessons, he adds. No. 1: don't be intimidated by scientists. (He's a lawyer.) No. 2: know when to say no. A few weeks ago, meeting with Stefánsson in Boston, he proved his point. The two were mulling the fate of deCODEme, the consumer diagnostic test. Stefánsson said he still hoped to be "very modestly marketing" the test as of next year. Collier raised his eyebrows and said, "If you want one, you'd better buy it now." In other words: forget it.
If deCODE's new, consulting-heavy business model can keep it afloat, the company will almost certainly continue to discover new genes. Even while it was in bankruptcy, it kept publishing. On Dec. 17, it revealed an intriguing new finding in Nature about a gene that behaves in an odd fashion—it raises a person's risk of diabetes if it's inherited from the father, but lowers the risk if it's inherited from the mother. The company was able to find that gene only because it had "smart-enough statistical techniques" to separate out which parent bequeathed it, says Randy Jirtle, who studies such genes at Duke University. Most geneticists don't have those techniques, so deCODE has an edge.
Of course, the firm's scientific success didn't save it from financial failure the first time around. And it's highly unusual for companies to come back from near death. But deCODE built itself up from very little before. Despite the odds, Stefánsson believes he'll be able to do that again.