My Beautiful Genome

by Lone Frank

  The young man sounds upset, but Stefánsson just brushes him off.

  “Those are the rules of the game. It’s in the nature of things that nobody can predict future findings, but we have to act at any point on the knowledge we have here and now.”

  He reminds me of the cancer researcher Bert Vogelstein, from Johns Hopkins University, who wrote something similar in Nature:“Humans are really good at being able to take a bit of knowledge and use it to great advantage. It’s important not to wait until we understand everything, because that’s going to be a long time away.”

  Aren’t we, perhaps, too impatient? A population concerned about health won’t tolerate waiting until everything is illuminated to the darkest corners.

  Nevertheless, the young man has a point. It is striking how many different messages you can get out of the same genome, not because our genes change but because tests and interpretations do. It is only mentioned in passing at the conference, but there are problems with the interpretations of the risks that accompany genetic profiles, because the providers make use of different gene chips in their analyses. Put simply, they do not test the same variants. Some test seven variants for cardiovascular disease, while others only take four into account, which means the same person can get the tag “high risk” from one shop and “low risk” from another. Some use only nine variants to calculate the risk for type 2 diabetes, while others use eighteen and, thus, are more exact.

  Then, there are the updates. Discoveries of significant new SNPs are incorporated with those already known and, just like that, the same genome provides a new risk assessment. In 2009, New Scientist asked a group of Dutch researchers to look more closely at the risk assessment for type 2 diabetes developed at deCODEme, and quite a bit had changed in just two years. The researchers did not test living people, but instead produced computer simulations of almost six thousand genomes with different combinations of SNPs related to diabetes – a “virtual patient” database. When deCODEme launched their gene profile in the autumn of 2007, eight SNPs in the gene TCFL2 formed the basis for its risk calculation, but just a year later this was upgraded to eleven, and in 2009, to fifteen. With these changes, four in ten computer-simulated people changed risk categories; one in ten changed categories twice.

  Yet would the experiences of virtual patients translate to people in the real world?

  Yes and no. You could say that changing risk categories doesn’t much matter if the disease in question cannot purposefully be prevented. But if you are talking about a condition for which preventative medical treatment can be recommended for one risk category but not for another, it immediately has great significance.

  Of course, consumer confidence is something else entirely. Cecile Janssen of Rotterdam’s Erasmus University, the researcher in charge of the study, admits she is worried that what there is might evaporate. If you expect that results, and your probable risk, will change over time no matter what you do, you lose the motivation to react to them. Like when the health authorities advise you to eat something or other to prevent cancer, only to withdraw that advice later.

  The skeptics say shut it down, it’s far too early, consumer genetics isn’t ready for prime time. No, no, no, others argue: the field must have room to develop if it is going to be useful, and it must take place in the spotlight of public scrutiny, where everyone can see what’s happening. Consumers must be prepared to accept uncertainty as a fundamental condition. We just have to get used to the fact that science does not always provide absolute, final answers; it is an eternally forward-moving process that constantly changes our understanding of the world.

  As Robert Cook Deegan, the director of Genome Ethics, Law, and Policy at Duke University, says from the podium: “No matter what the problems are now, we can’t get around the fact that the thinking in this area has itself changed fundamentally. There is access to personal genetic information, and the information keeps getting cheaper and better.”

  He calls the German law “pure idiocy.”

  “Wanting to protect people against consumer genetics is like wanting to shut down the Internet because there is porn out there that can corrupt unguarded souls. That’s what the Germans are doing with their prohibition. They are acting from fear and without any regard for the potential advantages they are missing out on.”

  Still, generally speaking, an interesting step has been taken. For many years now, a mantra has been chanted about “toxic knowledge” and the “right not to know.” The predominant attitude has been that you should ultimately protect “people” from knowledge that might make them uncomfortable, because as non-experts they are incapable of taking a position on it. A kind interpretation would label that as paternalistic guardianship. Or we could straight out call it the superciliousness of the experts. The first studies of how people react to genetic knowledge are only coming out now, and they indicate that ordinary consumers are extremely good at handling such “toxic” information about themselves.

  For years, Robert Green has been researching Alzheimer’s disease at Boston University. Among many projects, he and his colleagues have tested relatives of Alzheimer’s patients for the ApoE4 variant, which increases tenfold the risk of developing the disease. They found, quite surprisingly, that those who prove to have the – perhaps fateful – variant do not have a higher stress level and do not have a greater fear of the future than those who have not been tested and therefore do not know their risk.

  Admittedly, for the first six weeks after receiving the results of the test, the carriers of ApoE4 reported feeling more stressed than people without that certain knowledge, but a year and a half later, when the researchers spoke to them again, there was no difference in the psychological well-being of the two groups. On the contrary, some of those who knew they had the ApoE4 variant paid far more attention to planning their time and thinking about their lives than those who might have the same risk of Alzheimer’s disease but had decided not to learn their status.

  What then of the gene profiles and the range of disease risks associated with them? According to a study done by researchers at the Scripps Translational Science Institute, these tests do not seem to induce heightened anxiety. For six months, the team followed more than two thousand people who bought a SNP-based gene profile from the company Navigenics; there was no indication that the results caused the customers any distress. This prompted the leader of the study, Eric Topol, to tell the New York Times: “Up until now there’s been lots of speculation and what I’d call fear-mongering about the impact of these tests, but now we have the data.”

  Another interesting study, headed by Colleen McBride of the National Institutes of Health, investigated smokers and lung cancer. Smokers related to patients with lung cancer were offered a test for a gene variant known to increase the risk of this cancer in smokers. The hypothesis was that people who learn they do not have an increased risk of a given illness, or that they may even have a lower than average risk, will react with indifference. In other words, they will continue smoking if they think they have a genetic “free ride” when it comes to lung cancer and they will happily gobble down mountains of fattening fast food if their genes indicate the same for cardiovascular disease.

  For ethicists, this worry sounds plausible, but the research appears to undermine it. In an extension to the NIH study, no difference was detected in the motivation of those tested to stop smoking, whether or not they had the risk-increasing mutation. McBride sums up the two studies:“The findings may help us put a damper on these paternalistic concerns that we have to protect people from this type of information.”

  Paternalism has its dangers, too. Consider the disturbing story shared by Kári Stefánsson, in one of his panel discussions. In a study of genes associated with breast cancer, deCODE Genetics identified 110 Icelandic women who, based on the information contained in their family histories, it deduced had potentially lethal BRCA2 mutations, which give the carrier a seventy-five percent chance of developing breast cancer. This is
a piece of knowledge that can genuinely be called toxic, and the researchers passed it on to the country’s health authorities.

  So, a bunch of civil servants are sitting with this information that touches on the life or death of real people – some of whom they might even know personally. What do they do? Nothing. No one contacted the women. These BRCA2-positive women were walking around in ignorance, while their genes were bandied about at a conference at which top researchers mingled with representatives from personal care product manufacturers and health insurance companies. The women may be enjoying a happy ignorance, but wouldn’t you be seriously unhappy and disappointed with the health authorities if such information were withheld from you? I try to imagine the day they finally receive the news, and shiver in the cold auditorium.

  Once, the most pressing issue when conducting research on people was to ensure they were anonymous and remained so. Subjects volunteered information on the condition that they would never themselves benefit from it and that no one would provide them with their individual results. Now, this old record is sounding a little worn and particularly out of tune with the times.

  “I believe the most important ethical, legal, and social question to be studied in the field of genetics is the delivery of genetic results to the participants in the research,” writes Catherine A. McCarty of the Marshfield Clinic Research Foundation, in the Genomics Law Report blog. In the same forum, geneticist Daniel MacArthur of the UK Sanger Institute offers a hypothetical story of a woman who participates in a diabetes study, and the researchers find she has BRCA mutations. It would be crazy not to give her this information, right?

  THERE’S SOMETHING ELFLIKE about Linda Avey, daughter of a Lutheran minister from South Dakota, and one of the founders of 23andMe. It is not just her slender figure but her smile, which spreads effortlessly across her entire face, and her affable manner. You quickly feel she could be your close friend, someone to chat with, someone to confide in. Dark suits whiz along the corridors and, at one point, she leans in toward me and declares in a stage whisper, “Genetics is such a male-dominated world, isn’t it?”

  That’s hard to deny. The field seems a little pumped up on testosterone, bloated by male egos and their hi-tech toys. This may be the chief reason that 23andMe has aroused so much attention – the company was started by two women. Avey admits they are shaking things up by their very existence.

  She’s not shy about challenging the status quo. “We need a more up-to-date way of thinking about ethics. The research community has pursued the idea of ‘protecting human guinea pigs’ to the point of paternalism run amok,” she says. “We also need a whole new kind of research.”

  That is a grand ambition. I remind her that 23andMe has often been called genetics’s Facebook – and usually not admiringly. Consumers don’t just buy a gene profile from the California startup but also a gorgeous and user-friendly website, where they can invite other users of the 23andMe service to join in conversations and compare their ancestry and risk of disease. There is a special group for mothers-to-be, where they can keep one another updated on the progress of their pregnancy and compare symptoms and genetic information. Of course, some discuss whether their children should be gene-tested after birth.

  “The ‘Facebook’ format is quite deliberate,” says Avey, giving my arm a gentle pat. She explains that the choice was meant to encourage information-sharing among users and, in this way, create the basis for a greater change in people’s thinking about the research. “It’s not popular everywhere. There are a few, but very vocal, top researchers who seem to feel threatened by the very idea of democratizing DNA. Recently, I was at a meeting where one of them became so upset that he almost shouted at me, ‘You’re trivializing genetics!’”

  Avey bulges her eyes and shakes her head in a parody of the grumpy old man.

  “No, we’re not,” she continues, “but we are bringing genetics to the people, and we want to make research far more appealing to ordinary people. Because we believe there is a need for research in which the subjects are themselves actively engaged and they get personal knowledge out of the projects.”

  Whether or not Avey is proselytizing, this seems like an almost inevitable occurrence in today’s technological and social environment. First are the ever-cheaper tools for identifying genetic markers – from SNPs to gene sequences to the map of the entire genome. Then there is the fact that, in the age of Web 2.0, people are comfortable sharing even very personal data. Internet users no longer sit passively and get content shoved down their throats, but instead find it, share it, even create it themselves.

  “This idea was a part of our vision right from the beginning, and it grew out of my frustrations from many years in the medical industry,” Avey sighs. She looks into my eyes. “What is the biggest obstacle to medical progress? The challenge of all challenges?”

  For the moment, I can’t think of anything to say, so I smile knowingly, as if this were a rhetorical question. Avey continues, apparently without noticing my evasion.

  “It is choosing the right patients for each study and then to get enough of them signed up for the studies necessary to develop and test new medicine.”

  I nod. From working in biotechnology labs, one of which was a US-based startup in the field of neurodegenerative conditions such as Alzheimer’s and Parkinson’s diseases, I know that testing and research on people is colossal, resource-demanding work. Clinical trials, starting from tests on animals and working through many human phases to assess safety and efficacy, count for a huge portion of more than one billion dollars it costs to develop a new drug from test tube to market.

  Avey continues, a frustrated expression etching her face. “But the whole research culture is on the wrong track. You have all these university research scientists who feel like they own the disease they are researching.” White-smocked academics, who petulantly insist on my atrial fibrillation and my sclerosis. I’m going to solve the problem.

  “This sense of monopoly prevents the exchange of data, which slows everything down,” she says. Perhaps that is why, in September 2009, Avey chose to step down from 23andMe to launch the Brainstorm Research Foundation, which is devoted to collecting the phenotypes and health outcomes of people with genetic markers, including ApoE4, the Alzheimer’s disease marker that both she and her husband carry. “Having experienced this delay, it gradually became clear to me that the patients have to be at the center and that they should be the driving force in research and directly involved in it.”

  Lord knows, Avey’s message sounds warm and wonderful, but it’s not stopping the gentlemen in the hall behind us from chattering on about whether the public can handle taking a bite from the fruit of the genetic tree. Whether ordinary people will understand the uncertainty that follows from the fact that genetic data, in these early days, are far from exhaustive and subject to constant and intense development.

  “Uncertainty exists, and you have to be completely honest about that,” says Avey quickly and, perhaps, dismissively. “But listen: we see ourselves as creating a sort of ecosystem of patients and users. People who stay with us on the website, who keep up with developments and continue to enter in their data as they gradually get older. Can you see it? These groups – or cohorts – have built-in opportunities for conducting long-term studies that run for years. Studies that you cannot scrape together the money or the research subjects for today.”

  Yes, I can see it, and I even think it’s a good idea. You could follow the treatment people receive and discover unintended side effects of drugs that no one would otherwise have registered. Or you track the ways people live and uncover how they add up to differences in health, lifespan, maybe even life satisfaction. All these things would, of course, be evaluated through a genetic prism, understood through the connections between diseases, lifespan or other traits, and a person’s gene variants.

  “And,” says Avey, “everything is done on a voluntary basis and absolutely free of charge to the public health syst
em.”

  I begin to wonder about what this would mean for the user, the individual. Would participating in a genetics-based “Facebook ecology” be a way to engage more intimately with decisions about health and life in general? Time will tell.

  To start, 23andMe is tackling Parkinson’s disease. In a project supported and funded by a number of Parkinson’s foundations and patient associations, the company invited ten thousand patients to receive a gene profile almost free. In return, the patients volunteer to provide regular, detailed updates about their medical history. They are asked to relate what medicines they are taking and what effects and side effects they experience. But they also have to reveal how they are living in a broader sense: their diet, their physical exercise regime (or lack of), and a long list of other factors that researchers can compare to genetic markers to gain greater knowledge about the disease.

  “Mind you, we’re not just talking about information for researchers, and that is the point. This is all knowledge that the patients and their doctors can compare directly and use to optimize treatment for the individual. There is a need for more concrete knowledge – data, quite simply – in everyday medical practice.”

  On the heels of the Parkinson’s project, the company also started its so-called Research Revolution program, for which Avey and her cofounder Anne Wojcicki are hoping to recruit the vanguard from the consumers attending the genomics show (or reading newspaper articles about it).

  The Research Revolution – that sounds impressive; I have to admit that these two entrepreneurs are very good at turning a phrase. Their revolution – which will pave the way for Avey’s ecological system – is being presented as an essential component in the radical democratization of science, a sort of genuine research movement for the people. On the Research Revolution website, there are ten diseases to choose from; as a user, you help decide which of them you think 23andMe should tackle next. The illness that most quickly enrols a thousand volunteer patients will immediately get its own research project. You vote with a click and, at the same time, put your body to virtual service.