23andMe’s CEO on the Struggle to Get Over Regulatory Hurdles
Late on a Friday in November 2013 I was at a strategy offsite when my executive assistant texted that we had received a package by courier from the U.S. Food and Drug Administration. At the time, 23andMe had been in a years-long back-and-forth with the FDA over how we should be regulated, so this news made me anxious. Hoping to gain as much time as possible, I texted back, “Don’t sign for it!” She replied, “It’s too late, I already did.” As it turned out, that package contained a warning letter that would forever change the course of 23andMe.
On Monday morning the FDA released the letter to the press—something it rarely does with such short notice. Reporters began calling. Then David Kessler, the former FDA commissioner who’d been unofficially advising 23andMe since we launched, called. “Anne,” he said, “I know you’re probably not very worried about this letter, but let me tell you—you should be. The FDA is really angry.”
23andMe was then seven years old and provided the only direct-to-consumer genetic test with health-risk information. We’d been in discussions with regulators since the beginning but had usually been able to resolve any issues. This was clearly different. We were ordered to immediately stop offering health-risk reports. We suggested some changes in how we marketed the tests and offered to have them ordered by physicians, but the FDA said no, that wasn’t enough.
I’ve usually been good at navigating challenges, but this time I was stumped. I spent the next week calling lawyers, lobbyists, and scientific experts, and heard a variety of opinions. My initial reaction was to challenge the letter, but I soon realized that wasn’t the right move. I had one transformative meeting with the regulatory team at a large pharmaceutical company, whose members asked, “Do you want a short-term solution so that you can sell the company and be done, or do you want to stick with it for the next decade?” My answer was “I’m with 23andMe for the next decade and more.” “Well, then,” they said, “do the hard work and show the agency that this is a high-quality test that consumers can understand. That will have a far greater impact on the industry. But it will take you years.”
It was a painful transition and a significant cultural shift for the company. Our health-screening product stayed off the market for two years while we worked on FDA clearance and changed the core infrastructure of the company to be compliant. We hired new people and trained employees to operate within a regulated environment. We couldn’t take the traditional Silicon Valley approach of iterating quickly and launching; instead we implemented a compliance system with checkpoints to make sure we met all the necessary requirements.
Culture change is hard for any company, and we were unusual in that everyone eventually bought in. People often ask how we made it happen. There was no magic. I credit our chief legal and regulatory officer, Kathy Hibbs, and her team for leading that transformation and sticking with it. Change happens when strong leadership doesn’t waver with the message and gets employees to rally around the vision.
A Passion for Health
I grew up on the campus of Stanford University. My mother teaches journalism at Palo Alto High School, and my father is a particle physicist at Stanford. Together they taught me and my sisters to be passionate about what we do in life. That drove us to pursue jobs we love and to work hard at them. My sister Susan is now the CEO of YouTube, and my sister Janet is an associate adjunct professor of pediatrics at UCSF.
From an early age I have been interested in health, wellness, and the human body. I was very close to my pediatrician, Alan Bernstein, and used to love seeing him and peppering him with questions. My family had a copy of the Merck Manual, and I spent hours reading about diseases. By middle school I’d become fascinated with twin studies, which examine how environment and heredity combine to influence a person’s physiology and behaviors.
After high school I attended Yale, where I majored in biology. I loved molecular biology and considered pursuing an MD/PhD. But physicians I spoke with felt that the industry was changing for the worse, and there were more-interesting careers to pursue. I did a lot of lab research and enjoyed it, but I didn’t believe the PhD route was right for me. After graduation I worked on Wall Street as a health care analyst for a few years and then quit to travel, volunteer at a hospital, and prepare to take the entrance exam for medical school.
As I considered whether to apply, I got an opportunity to go back to finance at a hedge fund, where I invested broadly in health care: biotech, pharmaceuticals, medical devices, hospitals, nursing homes, and pharmacy-benefit managers. Investing was a great way for me to understand the health care sector while I figured out my next move. The longer I spent getting to know health care companies, the more I realized that the industry did not represent my true interests. I met all kinds of fabulous people who genuinely wanted to change health care, but the financial incentives of the overall system did not work to keep individuals healthy. It seemed wrong that no one made money if I stayed healthy but lots of companies would make money if I got sick.
Janet studies the obesity epidemic and can talk about the associated costs and the impact on lives. From a Wall Street perspective, I would talk about the opportunities to monetize all the comorbidities that come from obesity—and why the growing global diabetes epidemic was a reason to invest in the health care sector.
23andMe is the largest genetics community in the world for study.
I knew I needed to do something different. I spent my days thinking about how I could change the system, and I ultimately concluded that I don’t believe the system can change. If you want to change health care, you have to build from entirely outside the system.
I had always been intrigued by genetics while I was investing. Scientists had only recently mapped the human genome, and there was tremendous enthusiasm about what could come of it. I was doing research on a company called Affymetrix that had pioneered the first whole genome array—a low-cost way to look at a number of the known variations among humans.
I remember a conversation in late 2005 with a postdoctoral student at the Broad Institute, a genomic research center in Cambridge, Massachusetts, who said, “Anne, we’re on the brink of a revolution. We are about to solve everything.”
Later I had dinner with another genetics expert who was studying diabetes on a small island in the Pacific. Almost 100% of the island’s people were obese, but only 80% were diabetic. What was protecting the other 20%? The answer was most likely genetics. He told me, “I have so much data it’s chaos, but not enough data to make sense of it all.” If we had more data—the world’s health data—we could solve everything.
By then I was 32, living in San Francisco, and had just begun dating Sergey Brin, a cofounder of Google. I was surrounded by people who were starting up or working at companies that used data in interesting ways. Some of our friends were employed by social networking companies, and they schooled me in Web 2.0 and the power of social networks. That power in combination with low-cost genetic information suggested an opportunity to do something radical in research.
Around this time I met Linda Avey and Paul Cusenza. They had been talking about starting a company that would offer direct-to-consumer genetic testing. I joined them as a cofounder. After years of feeling frustrated with the health care system and powerless to do anything about it, I believed I was on a path to try something different and potentially revolutionary. We called the company 23andMe, after the 23 pairs of chromosomes that make up DNA.
A “Medical Device”?
We spent the first year traveling around, meeting with scientists, and assembling our advisory committee. We worked with a lot of experts on ethics, privacy, and legal issues. We felt a responsibility to get it right the first time, because we knew people would be fearful: What is more personal than your DNA? We built relationships with regulators.
The key question was whether our test should be considered a medical device—a category regulated by the FDA. A medical device is “intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease, in man or other animals.” It was unclear whether we met the definition. A wooden stick used in a popsicle is not a medical device. But used as a tongue depressor, it becomes one. Our genetic test told people about medical risks, but we didn’t intend it to diagnose or cure disease. So was it a medical device?
We discussed whether the container our customers use to collect and mail their saliva was a medical device, whether the test system was, and whether the information we report was subject to FDA regulation. We had several face-to-face meetings with FDA officials, and we aimed to be transparent and open to input. This was new territory for them as well, and Andrew von Eschenbach, then the FDA commissioner, indicated that the agency did not necessarily think our test was subject to regulation. We came away from those early discussions with the understanding that what we proposed did not require FDA premarket review.
Only years later did I realize how heavily politics influences the way an industry is regulated. During the George W. Bush administration the FDA had a mandate not to overregulate. Under President Obama the agency shifted toward more oversight. By 2010, after a competing company began selling DNA test kits at Walgreens, the Senate started holding hearings.
The new FDA commissioner made it clear that we would be subject to regulation. At a public meeting with the FDA in 2010 we said we would comply with the agency and would engage in figuring out how to be regulated. Regulators need input from companies to understand the nuances in the work they are doing. We helped the FDA craft standards and regulations that made sense. We believed we were making progress. But in retrospect it’s clear that we were far more out of alignment with the agency than we realized.
By 2013, 23andMe was doing well. We’d sold half a million DNA-testing kits and had received more than $126 million in venture funding. After years of slow growth, the market was taking off. Then, very suddenly, came that warning letter.
A Savvier Approach
Even now, I’m not sure what made the FDA change its view of 23andMe so dramatically and quickly. We had been promoting our product more aggressively with TV ads, at-home mailers, and some magazine covers. I don’t believe it was any one of those but rather the sum of all of them. Whatever the cause, the agency had lost its patience by the time we got the letter.
We were lucky that our product was used for two purposes: to help people learn about their ancestry and ethnic origins, and to understand what their genes might predict about their health. The FDA was forcing us to stop selling the health product, but we could continue to sell the ancestry product, and our customers could continue to get access to their uninterpreted DNA.
Our TV campaign had helped drive an unusually high number of orders—more than 200,000—in just a few months before the warning letter arrived. Our lab was backlogged with those samples. The FDA could have prohibited us from processing them and insisted that we offer refunds, which would have been disastrous for the company and forced us into bankruptcy. But it didn’t; we were allowed to return the health results to customers who had purchased prior to November 22, 2013. I took that as a sign that the agency wanted to work with us and believed our product had potential.
We learned from people with connections to the FDA that some officials felt very strongly that 23andMe should be reined in. We reached out to industry advisers who knew and understood the agency and had strong working relationships with some of its people. I wanted to start a conversation, but some folks did not even want to speak with us. The first time I emailed one adviser, she wrote back, “I am not a fan of 23andMe.” I kept emailing and listening to reasons that she (and others) had for disliking what we had done. Starting that dialogue with adversaries was important for me to understand just what expectations we would need to meet. It also demonstrated that our desire to be responsible and find a path forward was genuine.
Robert P. Calderwood (office); Jason Madara (tube, DNA model)
Clockwise from top: 23andMe headquarters, in Sunnyvale, California; a model of partially unwound DNA; the tube customers use to collect and mail their saliva for testing
Kathy Hibbs, whom we’d hired as chief legal counsel a few months after receiving the FDA letter, took the lead on working with the agency. Kathy had spent more than 10 years as general counsel with two other genetics companies and had also worked in the medical device industry. She was an ideal candidate. More important, she saw the potential for what we were trying to do with the direct-to-consumer approach and the research. She believed we could prove to the FDA that this was an accurate test that consumers could understand without the help of a physician or a genetics counselor. When she joined, it was a big shift for the company. Although the core mission was unchanged, we learned that we had to take a different approach to prove to the FDA that this was a safe product for consumers.
The FDA approach originally didn’t seem logical to me, but Kathy was able to break it down so that it made sense. With her in charge, I attended meetings on it only as needed. It was important for me to give her unwavering support but not to be the one leading the transition.
To get FDA authorization, we had to first agree on what the requirements would be.
Ultimately we focused on proving two things: that the test was valid, and that customers were capable of understanding the results we sent them.
Proving the validity of our test and data was generally straightforward and based on what other genetic products had submitted in their premarket reviews. The greater challenge was finding a way to demonstrate that the average U.S. consumer could clearly understand our results and how they should be interpreted. We conducted user-comprehension studies, surveying hundreds of people, of all ages, ethnicities, and educational backgrounds, and asking standardized questions to see whether they understood the information—not just that having a variant might put them at higher risk for a condition, but also that not having a variant didn’t mean they were risk-free. We have demonstrated repeatedly that our reports achieve at least 90% comprehension.
In February 2015, about 15 months after that warning letter, 23andMe received the first-ever FDA authorization for a direct-to-consumer genetic test to report carrier status, which informs people of variants that may not affect their health but could affect the health of their families in the future. Two years later we received the first-ever FDA authorization to issue genetic health-risk reports, which inform consumers of their personal risk for certain health conditions. Since then we’ve added conditions such as late-onset Alzheimer’s disease, Parkinson’s, and, following yet another FDA authorization, variants in the BRCA1/BRCA2 genes that signal increased risk for breast, ovarian, or prostate cancer. In the fall of 2018 the FDA granted us the only authorization to offer direct-to-consumer information on pharmacogenetics—how customers’ genes may influence the way they metabolize certain medications.
A Long-Term Vision
I tend to see a silver lining in everything. I definitely feel that way about our experience with the Food and Drug Administration. It made us a stronger company and created a better product that meets higher standards. The team that made it happen stepped up, learned what they didn’t know, and executed on the plan. I am most proud of our legacy of advocating for individuals to have direct access to the test. I believe that people of all education levels are capable of being in charge of their own health.
By being direct-to-consumer and affordable, we have enabled millions of people to learn about their DNA and their health risks, which has helped them take actions to prevent disease. We have also allowed millions of them to opt in to research, making 23andMe the largest genetics community in the world for study. We have published more than 150 papers and started a therapeutics team to create novel therapies based on genetic insights.
For instance, when the Covid-19 pandemic emerged, in early 2020, we quickly launched a study to determine whether any genetic factors are associated in those cases in which the virus is particularly severe. Within five weeks we’d enrolled almost half a million people, including several thousand with confirmed cases of the coronavirus. We aren’t studying only current 23andMe customers; we also sought out 10,000 other people who’d been hospitalized with Covid-19. It’s possible that we won’t find strong genetic associations for differences in the severity of symptoms. But we know from past research that genetics plays a role in both susceptibility to and severity of other infectious diseases, including malaria and norovirus. And depending on what we learn, our results could aid in assessing differences in risk among individuals and in ways of treating the disease in different patients.
It’s been 13 years since we started the company. Thousands of customers have reached out to tell us how 23andMe has changed their lives and, in some cases, saved them. Our mission—to help people understand and benefit from the human genome—continues to be our guiding light. I feel that we’re just getting started.
A version of this article appeared in the September–October 2020 issue of Harvard Business Review.