Whither Personal Genomics?

Posted by Jesse Reynolds January 29, 2010
Biopolitical Times
Two years ago, direct-to-consumer (or personal) genomics was the all the buzz. Media coverage was rampant, often the result of 23andMe's remarkable PR blitz of celebrity-laden "spit parties" in Davos and New York. Flash forward to the present: coverage is slim and the companies are suffering, including layoffs and one major bankruptcy. To where does the fledgling industry go from here? Three companies offer contrasting examples: The most visible one may be using a flawed business model and product. The oldest, and bankrupt, one has a troubling track record of privacy and consent. And a new one takes a new, and potentially problematic, approach.

Recent in-depth articles in The Times (UK) and The Washington Post explore the current status of the still-young industry. The former frames it:
Personal genetics was the next big thing: what celebrities and moguls were buying today, the masses would be buying tomorrow. 23andMe and rivals such as [bankrupt] deCODE Genetics, Navigenics and Pathway Genomics would do for DNA what Google and Amazon did for the internet. The genetic future was now.

Eighteen months on, that future stands postponed. Many doctors have rejected the hype about a new era of personalised medicine, arguing that most genetic screening is as likely to mislead as to inform. Neither have the public taken to consumer genomics: 23andMe, the market leader, has signed up only 30,000 subscribers in two years and has endured two rounds of redundancies....

Have science entrepreneurs such as [23andMe's Anne] Wojcicki and Kári Stefánsson, deCODE's founder and chief executive, jumped into a market for which the world is not ready? Will their companies prove to be the genome's Googles and Amazons, or its Boo.com?
I've long been of the mind that, just as the traditional business model of newspapers is to get revenue not from readers but from advertisers, personal genomics companies see the potential profit not from the consumers themselves but from the compiled databases--likely in the form of selling access to them. One key question is, just how reliable is this data? For example, 23andMe uses consumers' self-reporting:
But there is disagreement about the scientific worth of such a database. Stefánsson thinks [23andMe's data] will be useless because self-reported traits are unreliable and customers will object to the release of information. However, 23andMe says that the plan is already working and has led to the discovery of genetic variants linked to traits such as the curliness of hair. If this approach can be repeated with, say, responses to drugs, it's easy to see how a pharmaceutical company might bite.
This strikes me as wishful thinking at best. Consumers can likely self-report the curliness of their hair with significant accuracy. But I am skeptical about their ability to do the same for reactions to drugs, considering, for example, the remarkable (and rising) effectiveness of placebos.

What interests me more is the fate of the databases. As pressures to turn a profit mount during the Great Recession, and as companies such as deCODE, which recently filed for bankruptcy protection, are saved by new investors and subsequently restructured, what becomes of the data?
23andMe insists it will never release identifiable information to a third party without the explicit consent of the person concerned. "I'm always careful when I say we're going to have a revenue stream from the database," Wojcicki says. "We will not sell individuals' data, but the database as a whole will have enormous value."
Explicit, informed consent when purchasing a product online is difficult to truly obtain. When was the last time you read a user agreement when installing software?

The case of deCODE is even more troublesome, and not just because of its restructuring. At the Genomics Law Report blog, Dan Vorhous notes:
Stefansson and deCODE have been adamant that the change in ownership will not affect how the company uses data from customers of its deCODEme service, or the security of that data, but this is an issue that will continue to bear watching. As I’ve written elsewhere, deCODE’s new owners remain (legally) free to alter or expand their use of genetic data within a range of allowable uses.

The company’s announcement also refers in several places to genomic sequencing, and as Jocelyn Kasier first reported last fall, deCODE is planning to sequence the complete genomes of 2500 individuals from its Icelandic database by mid-2011 as it continues to search for the rare variants that may contribute some of the so-called “missing heritability” to common diseases and traits. According to Stefansson, deCODE “will not need to recontact these individuals for consent because their original consent agreements cover whole genome sequencing.”

As Stefansson and others continue to note, genomic research has already begun the transition from genotyping to whole-genome sequencing. Moving from examining a handful – even thousands – of an individual’s genetic markers to the sequencing of his or her entire genome creates the potential to understand that individual in much greater detail. It also carries with it a new and expanded set of considerations and risks that should impact any informed consent process....

What I do know is that deCODE has a history of aggressively interpreting when and where individualized informed consent is not required. The failed Icelandic Health Sector Database, which deCODE was instrumental in designing, relied on the now-discredited principle of “presumed consent.”
Just today, The New York Times profiled Counsyl, a new entrant into the field that is just now moving into the public eye. It explicitly targets prospective parents, arguing that preconception genetic testing can prevent the transmission of some diseases. In fact, the company (among whose advisors and investors is Henry Louis Gates (1,2)) frames their work as a "Campaign to End Preventable Genetic Disease."

To their defense, Counsyl emphasizes and currently tests for fatal and serious genetic diseases (although I am admittedly unfamiliar with many of these rare conditions). Furthermore, it backs up its argument that genetic testing is a right by offering financial aid for those who can't afford the $350 test, emphasizing that "Underserved groups and families with a history of genetic disease will receive priority."

But on the other hand, Counsyl makes a couple of significant misrepresentations. Its website says "Everyone is supposed to get genetic testing before pregnancy," even though the American College of Medical Genetics (or ACMG, who is cited) makes more cautious recommendations [PDF] about a limited number of diseases. Moreover, the ACMG recommends [PDF] that "A knowledgeable professional should be involved in the process of ordering and interpreting a [direct-to-consumer] genetic test." Finally, as the Times article notes, "The company calls its product the Universal Genetic Test, for example, even though there are thousands of genetic diseases, not just the 100 Counsyl tests detect."

More troubling is the eugenic potential of a direct-to-consumer genetic test aimed at potential parents. Although Counsyl currently emphasizes serious diseases, it is foreseeable that it will seek to expand its customer base by adding less serious--and even nonmedical--conditions. Moreover, its financial interests align well with those of the $3 billion baby business:
The test is already offered by more than 100 fertility clinics around the country...

Fertility centers have an economic incentive to promote the test because one of the best ways to avoid having a baby with a genetic disease is to test embryos created by test-tube fertilization before implanting them. This pre-implantation genetic diagnosis costs $3,000 or more, in addition to the roughly $10,000 fee for in vitro fertilization.

Notably, Counsyl's website recommends preimplantation genetic diagnosis and donor gametes, but not adoption, as responses to a double-carrier test result.

23andMe may have a defective product, despite its publicity. And deCODEme may raise serious privacy concerns. But Counsyl may represent the next step toward a consumer driven, technologically based, new eugenics.

Previously on Biopolitical Times: