After the biotechnology startup 23andMe ran afoul of the FDA last year for not properly backing up health claims it was making based on customers’ genetic results, 23andMe’s primary business seemed doomed. Indeed, until it can iron things out with the federal agency, the company is limited in the U.S. to marketing only its DNA ancestry kits to consumers.
But the controversy hasn’t stopped the company from doing what it set out to do when it launched in 2006: provide customers with a better understanding of their bodies and the DNA they’ll pass to their children, and offer up all of that data to science—for a profit. In recent months, 23andMe has brought its popular Personal Genome Service to customers in Canada and the United Kingdom, giving them an opportunity that’s currently off limits in the States: fill a tube with saliva, mail it back, and have it analyzed to determine some 100 risk factors for a number of diseases, inherited conditions, and adverse responses to certain drug therapies.
That service, along with its ancestry service, powers the B2B aspect of 23andMe’s business: users can opt in to share their genetic information, and 23andMe shares or sells that data to a range of partners—from academic labs to pharmaceutical companies—in an arrangement it says can ultimately help humanity as a whole.
Last week, the company announced a reported $60 million deal with Genentech to generate whole genome sequencing data for members of 23andMe’s Parkinson’s disease community. This week, it revealed that it will give drug maker Pfizer access to 23andMe’s genetic database. (The company is also waiting for FDA officials to approve just a single test as a “medical device”: a DNA analysis for Bloom syndrome, an inherited disorder that often results in deadly cancer by the mid-twenties.)
While the company doesn’t disclose the specifics of its data sharing agreements, they have always been central to 23andMe’s strategy. “The long game here is not to make money selling kits, although the kits are essential to get the base level data,” Patrick Chung, a 23andMe board member, told Fast Company last year. “Once you have the data, [the company] does actually become the Google of personalized health care.”
Ironically, as 23andMe works to resolve its issues with the FDA for its health products, its data-analysis initiative is earning support from other U.S. government agencies. In July, the company received a $1.4 million grant from the National Institute of Health to help expand its genotype database. A study released last week by one of the company’s researchers that illustrates the geographic spread of race and ethnicity throughout the U.S., based on the anonymized data of 160,000 customers, was funded in part by the NIH and the National Science Foundation. Among its findings: About 3.5% of self-identified European Americans have at least 1% or more African Ancestry, and one in every 20 African-Americans have Native American ancestry.
Just as apps like Mint and LinkedIn have transformed how we access, understand, and manage our financial and professional data, 23andMe, via its Personal Genome Service, wants to become a central storehouse for our genetic data. Founded by Anne Wojcicki, who is married to (though separated from) Google cofounder Sergey Brin, a carrier of Parkinson’s disease, the company—named for the 23 pairs of chromosomes in a normal human cell—wants to offer users the tools they need to understand their genetic information, and how it might relate to their health. And they see the timing as right, as health care slowly but surely moves from the generic one-size-fits-all approach to more unique treatments based on your genes.
Users appear interested in sharing their genetic data with the world, provided it’s anonymized. The company says that over 80% of its users have opted in to its data-sharing system, tantalized by the hope of contributing to important science that could eventually come back to help them and their families. As one 23andMe user who found out she was predisposed to breast cancer told me, “Genetic data is the most personal data I own, but if my data can contribute to finding better treatment or even a cure, why should I think twice about sharing it?”
By effectively “crowdsourcing” people’s DNA data, big pharmaceutical companies can do research that even their own large budgets might not otherwise allow. “Traditional research can take more than a decade and millions of dollars to conduct studies with just under a few hundred participants,” says Angela Calman-Wonson, VP of Communications at 23andMe. “We can undertake real-time research initiatives drawn from the more than 600,000 23andMe customers who have proactively elected to share their de-identified genetic information for research, and answer survey questions. This approach eliminates recruitment times, minimizes cost, and reduces the amount of time it takes to conduct research.”
It’s an approach that has caught the attention of academics and, increasingly, large pharmaceutical and biotech companies just like Pfizer, which had worked with 23andMe before this week’s announcement, to study the genetics of inflammatory bowel disease (IBD), and Genentech, which had previously paid the company to help it find breast cancer patients who had taken its drug Avastin. The approach has caught the attention of funders, too–so far, the company has raised $126 million. 23andMe won’t reveal how much it charges partners, which include the likes of The Michael J. Fox Foundation–for access to anonymized user data, but the company says it has more than 30 active collaborations with academic research centers and industry from around the world.
Those partnerships, over the past four years, have led 23andMe to publish or contribute to 22 peer-reviewed papers about a range of areas, including allergies, asthma, hypothyroidism, myopia, and breast cancer. These kinds of contributions, the company says, are the direct results of 23andMe’s users’ willingness to share their anonymized genetic data.
“For example, researchers came to 23andMe to learn whether or not a certain gene was more prevalent in cancer patients,” says Calman-Wonson. “23andMe sent surveys to individuals in our database with that particular gene, asking several cancer-related questions. We received more than 10,000 responses in 12 hours, and were able to determine that the gene was not prevalent among cancer patients. This type of research typically takes months and thousands of dollars, and in this case it took about 48 hours.”
By scaling up, she says, 23andMe could spark greater price drops in genomics—and even bigger follow-on effects. “We are really just scratching the surface of what is possible,” she says. With 23andMe, “we believe we can achieve the scale necessary for breakthroughs much more quickly and efficiently than traditional research methods.”
There are causes for concern, say critics of genomic testing, which tend to revolve around the concept of confidentiality. At Scientific American, science journalist Charles Seife summed up the worries by calling 23andMe’s Personal Genome Service “much more than a medical device; it is a one-way portal into a world where corporations have access to the innermost contents of your cells and where insurers and pharmaceutical firms and marketers might know more about your body than you know yourself.”
Still, on its website, the company explains that its data-mining analysis “does not constitute research on human subjects”—which is to say, it’s not subject to the rules and regulations that typically protect experimental subjects’ welfare and privacy. It also offers a warning: “Genetic Information that you share with others could be used against your interests. You should be careful about sharing your Genetic Information with others.”
Calman-Wonson also says users can change their mind and opt out of the research program at any time or, if they choose, can close their account at any point, and their data will be deleted from the company’s systems within 30 days. Most users, however, are proactive about sharing their data, she says.
“To date, more than 80% of our 800,000-plus customers have opted in to our research, and most answer survey questions,” says Calman-Wonson. “A large number are very actively engaged in answering questions every month. On average, each customer who opts in to participate in research informs more than 230 genetic studies.”
One portion of 23andMe’s user base constitutes the world’s largest group of Parkinson’s patients participating in research in the world.
“For our Parkinson’s research community, we were able to enroll 3,400 Parkinson’s patients, identify two new genetic associations for the disease, and publish the findings–all within 18 months. Today our Parkinson’s community has more than 10,000 people participating,” says Calman-Wonson.
Still, it’s not entirely clear how the research side of 23andMe’s business can continue to grow given the current hold on its Personal Genomics Service. Calman-Wonson admits that before DNA analysis and genome-based medical treatments can become commonplace, more research is needed—including the kind of research to which 23andMe is contributing.
“We are still operating in a comparatively new field,” she says. “It has only been about 10 years since the human genome was first mapped, and we still have so much to learn about how DNA works. The human body is incredibly complex, and genetics are just one factor of many that informs human health.”