Patently unclear – 23andme’s position on DNA patenting

Today I have been preparing a talk about personal genomics for a workshop in Edinburgh next month where I plan to discuss the tension between democratisation and commodification in the nascent consumer genomics sector. In a rather haphazard  bit of Google searching I was rather surprised to come across a patent application filed in November 2010 by the US consumer genomics firm 23andme. The patent relates to genes associated with Parkinson’s Disease, or to be more precise, and to quote from the patent application:

“The present invention is related to polymorphisms associated with Parkinson’s disease. More specifically, the invention is related to compositions, methods and for use in therapeutic and preventative treatment, study, diagnosis and prognosis of Parkinson’s disease.”

The diagnostic portion of the patent would appear to be fairly broad, covering risk prediction, diagnosis and prognosis using SNP analysis or gene expression data. (I am neither a scientist nor a patent lawyer and would welcome comment from those more qualified on the precise scope of the application). The current application supercedes earlier applications filed by the company in November 2009 and June 2010. Followers of the company will recall that 23andme had announced the discovery of new genes associated with Parkinson’s Disease having undertaken their first clinical research study in 2009-10, and the patent application would appear to be concerned with these novel discoveries.

The patent again raises two questions which I have long pondered: can you make money out of consumer genomics, and if so how, do you do it? Not long after launching their company in 2007, 23andme CEO and co-founder Anne Wojcicki was
quoted on the VentureBeat blog
expressing uncertainty about their business model:

“It’s really too early to specify how we might monetize and derive value from the information we’re aggregating. We’ve thought about a lot of different ways to monetize it, but we’re not ready to talk about them”

One division within the nascent consumer genomics industry has been on the issue of intellectual property derived from original research. Some companies, like deCODE, have spent years (and much money) on research uncovering new gene-disease associations which they have then patented. Others, like Navigenics, have trawled the scientific literature sifting through other people’s discoveries for genes which it might be useful to test for. 23andme seemed to be more in this camp – they talked a lot about the research value of their embryonic customer biobank when they launched, but it was unclear what they were going to do with the data, and the vast majority of the gene-disease associations they report to their customers have come from the scientific literature. The company’s decision to begin testing for the BRCA 1/2 genes in 2009, in apparent contravention of Myriad Genetics fiercely defended patents on the genes, suggested that their philosophy was provocatively open source in its attitude to the human genome.

A hostility to gene patents was certainly expressed by Linda Avey, one of the company’s founders, when we talked after a meeting at the European Parliament in Brussels in March 2009. Avey subsequently commented publicly on the topic in her blog a couple of times the following year (after her departure from 23andme). After attending an industry meeting in March 2010 she expressed her distaste for Myriad Genetics in no uncertain terms, saying their business model “made my skin crawl … the thought that a company is profiting so handsomely from intellectual property that corners the market on information simply residing in the human genome is unfathomable “ Unsurprisingly, Avey saw a fundamental conflict between Myriad’s approach and the business model of the company she had co-founded: “As individuals are able to access their entire genetic profile, and all the increasingly pertinent health information therein, the math just won’t work if gene variants, along with the inherent knowledge regarding disease or pharmacogenetic associations, are patented.”

A few months later on Dan Macarthur’s Genetic Futures blog, Avey commented on the 2010 ruling against Myriad’s patents:

“My hope is that this ruling stands and companies will need to actually innovate and create new advances BASED on genetic findings, not dependent on sole access to them. Rather than relying on obscure patent language and legal strategies, companies will need to develop products that are competitively positioned, something that will be good for consumers i.e. patients. With all due respect to the legal profession, we need to have more scientists and innovators in the lab, not lawyers.”

Given the apparent hostility to gene patents expressed by one of the company’s founders, 23andme’s Parkinson’s Disease patent application raises a number of questions, not least did Linda Avey resign in protest because of plans to file the first application in 2009? Had the company always intended to patent its gene-disease discoveries, and if so why was it not explicit about its plans? Given the company’s avowed mission to ‘democratise’ genomics, what were the participants in the Parkinson’s Disease study told about the intended commercial exploitation of discoveries arising from the study, and did anyone ask them what their preferences were?

Of course it may well be that 23andme is in fact committed to open source genomics and their patent application is simply designed to prevent other less well-intentioned actors from patenting their discoveries and then blocking their free use. But in the absence of public discussion of this topic by the company we are left with uncertainty as to their intent. 23andme proudly proclaim that their mission is “to be the world’s trusted source of personal genetic information.” That ambition sits uneasily with the lack of transparency in the company’s approach to intellectual property rights. In 2007 the company was not ready to talk publicly about the ways in which they might derive commercial value from their customer database, five years later that discussion is long overdue.