Amgen Buys DeCODE – Reflections Backwards, Forwards, and on DTC Genomics

Earlier this week, U.S. biotech giant Amgen announced it was buying DeCODE Genetics for $415 million, thus apparently ending that company’s “interesting” 16 year ride. See this article in Genome Web Daily (access requires free registration).

The notably abrasive (and I’m trying to be nice) Icelandic geneticist Kari Stefansson founded deCODE in 1996, in part to take advantage of Iceland’s small (about 280,000 people), genetic homogenous, and well documented (both for health and for genealogy) population for genetics research.  It made headlines in the late 1990s with its proposal to create a universal health database of Icelanders for research purposes.   The proposal was controversial, leading to the creation of an Icelandic group for research subjects’ rights, Mannvernd, to battle it, but the Icelandic parliament, the Althing (the oldest existing parliament in the world, it claims), passed the proposal in December 1998.  In January 2000 the Icelandic government formally awarded the licence to run this Health Sector Database, as expected, to deCODE.  deCODE issued stock privately, both after the legislation passed and again after the license was awarded.  Many of these shares ended up in the hands of Icelandic banks, other institutions, and individuals.  It also struck an apparently lucrative deal with international pharmaceutical company, Roche.

Contrary to what is widely believed, deCODE was not given any special right to collect DNA from Icelanders – just to build a database of their health records that could be used for (presumably lucrative) research purposes.  I wrote a fairly long article about deCODE through this point and published it in 2000:  Henry T. Greely, Iceland’s Plan for Genomics Research: Facts and Implications, 40 Jurimetrics J. 153–191 (2000). (Unfortunately, the article does not seem to be available on-line.  I’ll try to remedy that.)

What happened after my article came out was also interesting.   Based in large part on plans for its use of the Health Sector Database, in June 2000 deCODE made its first public offering, selling shares at over $18 that were quickly bid up to over $30.  But deCODE never built the database.  Citing, I suspect disingenuously, an Icelandic court decision against including the records of dead Icelanders in the database, it gave up any idea of spending the hundreds of millions of dollars that database would cost.  Instead, it turned its attention to disease gene discovery the old-fashioned way, collecting DNA and health information voluntarily from Icelanders with family histories of particular diseases.  It had some scientific successes over the next few years, but kept losing money.

A few years later, company was delisted from NASDAQ and, in November 2009, filed for bankruptcy under Chapter 11 of the U.S. bankruptcy code.  The bankruptcy listed assets of just under $70 million and debts of more than $313 million.  The shareholders, including many individual Icelanders who bought stock enthusiastically at high prices, were wiped out.  A few months later, the assets of deCODE were bought by a newly organized company, Saga Investments, that retained Stefansson in management.  A “new deCODE” was launched, as a private company (i.e., without publicly traded stock) that continued its gene discovery researc.  The new deCODE also continued deCODEme, a direct to consumer genomics service it had started in late 2007, before the bankruptcy. At that point, at least three companies were offering direct to consumer genomic services:  deCODEme, 23 and Me, and Navigenics.

Amgen announced that it had no plans to continue deCODEme.  This was not surprising; deCODEme never lowered its prices to meet competition from 23andMe.  It was charging nearly $1000 for services similar to those 23andMe was providing for under $300 (or less with frequent “sales”).  By the end it is unclear whether deCODEme had any customers.

So, what reflections?  Well, first, looking back, I fear the shareholders who were wiped out in the first deCODE bankruptcy will not see a penny of this week’s $415 million sales price.  And the Icelanders, about 140,000 of them deCODE has said, who volunteered their DNA and health records to deCODE are also not going to see any of that $415 million.  I suppose, perhaps, they can take some comfort in knowing what their contribution is worth – some substantial chunk of $415 million (potentially as much as $3,000 per person, though, practically, probably about a quarter to half of that) but only to the last owners of deCODE.  deCODE rode its Health Sector Database license to high stock sales (including shares owned by the founders).  Insiders did well out of the first deCODE; the little guys, no so much.  And at least some of those insiders were part of the second deCODE and presumably also did well from its sale.  This seems to me at least one more example of life’s unfairness, whether or not inevitable.

Looking forward, there is now only one major player for DTC genomic services.  deCODEme is gone.  Navigenics disappeared in July 2012, when it was purchased by Life Technologies, a San Diego area making of sequencing machines.  Life Technologies announced that it was not continuing the DTC business.  Only 23andMe survives – and that privately-held company (with deep-pocketed investors, including both Google and Google founders as individuals)  just announced, also earlier this week, that it had raised $50 million in new funds.  It also announced that because of the new financing, it was dropping its price to $99 to attract more customers.  See the Genome Web Daily report here. There are some smaller players in the DTC genomics world – according to the Genome Web Daily report, a company called Gene by Gene just announced its intent to do whole genome sequencing as a DTC service, though only providing raw data and not medical interpretation.

That limitation probably is a result of continuing FDA interest in regulating DTC genetics and genomics companies.  That interest first surfaced in May 2010 when Pathway Genomics attempted to offer DTC genomics at Walgreen drug stores – see the extensive coverage, over the years, of the FDA’s actions and interest in the Genomics Law Report (run by Dan Vorhaus, who really should be a law professor).  You can find the GLR here, along with, I now see, its own, typically excellent, take on the deCODE/Amgen deal (which I just now spotted), here.

What does all this mean?  Well, DTC genomics has not, so far, been a very profitable field. Whether 23andMe, with its deep pockets, will see it through to profitability is unclear. Although it now has no real competitors to worry about, it is lowering its prices, not raising them. The current DTC genomics model, using SNP chips, will certainly be obsolete soon (at least for most purposes), to be succeeded by whole exome sequencing (already offered by 23andMe for just under $1000) and, eventually, whole genome sequencing.  What that market will look like, who will be the main players in it, and how the FDA will affect it all remain to be seen.

In the meantime, goodbye deCODE.  According to the Genomics Law Reporter, Kari Stefansson may end up surviving yet another corporate change in deCODE to play a managerial role at of the former deCODE assets at Amgen, so it is not goodbye Kari.  I suspect, for better or for worse, we have not heard the end of him yet.

Hank Greely

3 Responses to Amgen Buys DeCODE – Reflections Backwards, Forwards, and on DTC Genomics
  1. I’m nostalgic. I began my molecular bio training right when deCODE was getting off the ground. It was promised as a hope to all humanity; a source of cures for everything. That also about the same time Wired came out with its “long boom” article (http://goo.gl/kxcfG) arguing that this thing called “the Internet” was going to ensure 25 years of sustained growth. The world was looking up, up, up.

    Its business failings aside, deCODE had some truly amazing discoveries, not the least of which was a common 900kb inversion that had, somehow, evaded everyone else for decades. It did some remarkable and useful sequencing work, when such work was thought to be incredibly important, patented little of it relative to its competitors, and published it widely and often.

    But, like “The Long Boom”–which, now, can be mocked mercilessly–its clear it wasn’t meant to be. All that sequencing work has translated into little, and deCODE contributed to the Icelandic economy’s recent fallout.

    Sometimes, promises of future success aren’t enough, and we’re left, empty-handed, to wistfully look back on the days when our former hopes gave us substance. Just like when Soundgarden broke up.

  2. Jake, I’m guessing you never saw Kari Stefansson talk – or knew, at the time, the dark side of how deCODE exploited Iceland and Icelanders, or, to be fair, how Iceland and Icelanders largely lined up to be exploited. To me, your remembered golden days seem made of pyrite.

    deCODE was no Soundgarden (whatever that was!)

  3. Pingback: Perception management | Thor's Web

Comments are closed.