by Lone Frank
Some might see it as a kind of magical thinking that I stay far away from any actual examination. I never fondle myself to search for lumps and, when anyone asks whether I shouldn’t have regular mammography, I usually mumble some defensive witticism as “they can’t do that for such small breasts.”
Now I’m sitting in front of a bright but cold screen that says that it can reveal the risk carried in my genome as defined by eight variations on chromosomes 2, 5, 8, 10, 11, 15, and 16. For women of Caucasian descent, the average risk of developing breast cancer is around twelve percent. Twelve out of a hundred ordinary women will get it at some point – no way around it. But depending on the combination of these seven variants, you can come up into the vicinity of a sixty percent risk. I feel certain that I am one of the six, and keep my eyes closed long after I’ve clicked.
But it’s unbelievable: I have a lower than average risk – not even the ordinary twelve percent, only 7.7. It is as if a very old, hissing pressure deep inside my body quickly seeps out and floats away.
For a wonderful, absurd second, I believe I’ve been liberated from death.
“WOW. YOU LOOK like you could use a strong cup of coffee.”
It’s seven o’clock the next morning, when deCODEme’s head of public relations, Edward Farmer, comes to retrieve me from the hotel. It’s not just cold in Iceland, it’s also dark, and to boot, I haven’t slept well. First, I had to celebrate my great relief at my lowered risk of cancer with a couple more beers, and later, when the euphoria wore off, I conscientiously studied my whole gene profile, one end to the other. From atrial fibrillation and baldness to gout and restless legs syndrome – all delightfully low. Sure, baldness was a given, but I wasn’t going to deny myself the delight.
Still, my night’s rest was uneasy. I wound up wondering how usable these genetic risk assessments really are, when it comes right down to it. Not just for me, but generally. What do you do with the information that you have a slightly smaller risk of getting asthma than the average? And what does it mean for your everyday life that your risk for kidney stones is a bit higher? What would I have done if my gene profile had given me a sixty percent chance of getting breast cancer? And how sure are they, anyway, about these molecular risk assessments?
“Put your seat belt on,” says Farmer, “I’m sure Kári will be able to explain all that.”
By Kári he means Kári Stefánsson, deCODEme’s founder, director, public face, and a bit of a story unto himself. The last time I had the pleasure was over ten years ago, when deCODE Genetics – the parent of deCODEme – was an infant and a hot topic of debate. The company was among the first to latch on to industrial genetic development, which was just beginning to shake things up. Stefánsson had a good idea, and he planned to realize it by drilling into the genome of the Icelandic people. He imagined a gigantic database in which the medical records of living Icelanders would be collected and combined with their genetic information and the genealogical records of the population going back centuries. Here was an efficient way to trace the genes that make us sick.
Stefánsson scraped together twelve million dollars from a group of believing venture capitalists, left his pleasant and prestigious position as a professor of neurology and neuropathology at Harvard University, and went home to Iceland, where he began to steamroll his will through the legislature. There was bitter local opposition from those who thought the genomes of a population should not be put into the hands of a private business, and the outrage spread far beyond the borders of this tiny island nation. Stefánsson became a sort of itinerant verbal prizefighter, famous in the media for his predilection for dark Armani suits and for his wrathful, Viking-like appearance. He went from conference to conference to address his critics and quickly gained notoriety for his intrinsic stubbornness.
Now in his sixties and majestic at six feet tall, with broad shoulders, white hair, and a beard, Stefánsson is, for good measure, the descendant of Iceland’s beloved poet and national bard of the tenth century, Egill Skallagrímsson. Ironically, Egill was not only known for his unusually beautiful verse but also for his unusually horrible behavior. If the food and drink were not to his like, the great poet might vomit on his host and, if anyone offended him, he did not shrink from gouging out their eyes or maiming them in some other way. Today, some of the labels tacked onto his descendant are “boastful,” “aggressive,” and “notoriously gruff.”
On deCODEme’s homepage, however, Stefánsson is portrayed with a warm smile, his chin resting thoughtfully in his hands. A winsome grandfather, were it not for the tight, black T-shirt that emphasizes his buff upper arms, and the text beneath the portrait:
Unlike a large number of other companies hoping to follow on our heels, deCODEme is not just a website that happens to be about human genetics, it is your portal into the world’s largest and most successful effort to understand the inherited risk of the most common diseases in contemporary society.
Arrogantly put, but essentially correct.
Stefánsson’s concept has been proved to hold water. For ten years now, deCODE Genetics has ransacked thousands of Icelandic genomes, from which scientists have identified an impressive battery of gene variants that affect human health and the risk of disease. It regularly spits out articles for the best scientific journals. Farmer has been kind enough to supply me with a pile of the most recent.
Professionals are impressed by the research but not everyone thinks it is a good idea to offer gene profiles directly to consumers, as the company now does. The skeptics include, among others, the UK Human Genetics Commission, which is worried that ordinary people cannot really understand the information without thorough, professional, genetic advice. The commission’s members argue that tests simply stating a risk should not be sold directly to consumers. Many doctors are of the same opinion and, in 2008, pressure from physician interest groups prodded the State of California to prohibit businesses from performing genetic tests that were not ordered by a licensed medical professional.
That seems to be the trend. In 2009, after years of economic difficulties, deCODE Genetics did a U-turn and attempted to reinvent itself as a “diagnostic company.” It filed for bankruptcy and was saved at the last minute by a group of American investors. Soon there will be money – lots of money – to be earned on gene profiles. As Stefánsson says in the press release that announced the decision: “As the focus of our healthcare system shifts toward prevention, measuring and controlling the individual risk of disease will become a central part of everyday medicine.”
IT HAS BEEN ten years since my last meeting with Stefánsson, but his greeting tells me that he has not changed much.
“Are you really going to burden me with your company once again?” he says, extending his hand without rising from his chair or looking up from his papers. Before I can answer, he shouts to his assistant in reception that he needs some coffee if he is going to survive this chat. A classic tack from his repertoire; one a number of my journalist colleagues have written about. What surprises me is not his rudeness but his voice – I had completely forgotten how pleasant it is. Soft, as if wrapped in fur. His English has a thick accent that is really quite charming, even now, as he tells Farmer that as head of communications he can be present during the interview but should be prepared to keep his mouth shut.
“Completely shut.”
I figure that the only way to avoid being completely shut out is to make the first strike. I begin by poking an old wound. Stefánsson’s startup was the first to market the personal gene profile, but its competitor, 23andMe, based in Silicon Valley, is far more well-known, and presumably more commercially successful. I ask with an air of innocence what Stefánsson thinks about the fact that 23andMe featured prominently when Time magazine crowned the gene profile as “invention of the year.” A brief but very cool look greets me from the other side of the desk.
“It did not surprise me, not at all, and I’ll tell you why. One of its two founders shares a bed with a rich
, famous man, and that’s the sort of thing Americans fall for,” he growls. “Okay, I was pissed off for an afternoon. But so what? I’ve been pissed off a lot of afternoons in my life.” He looks directly at Farmer, who commendably maintains his PR mask of neutrality.
I recall that Anne Wojcicki of 23andMe – the woman who “shares a bed” with Google’s founder Sergey Brin – made a very successful appearance, while visibly pregnant, on the Oprah Winfrey Show. Somewhat later Stefánsson was left to explain the blessings of gene profiles to the lower-rated Martha Stewart, the ever-smiling goddess of domesticity whose stint in prison for securities fraud was then still recent news.
“Tell me … who is the typical customer for deCODEme? An average viewer of the Martha Stewart Show?” I ask, hoping to sound innocent.
“Where the hell did you find this one?” says Stefánsson, apparently to Farmer, but this time his growl is accompanied by something resembling the early stages of a smile. The ice is broken, and we can finally start the conversation in earnest. The great man shifts in his chair, turns for the first time in my direction, and explains that the company sells most of its gene profiles to professionals.
“General practitioners, but also clinics that specialize in prevention, which are shooting up all over the US,” he explains. “The number of private customers is increasing, of course, but in the long term the healthcare system is the biggest market. In fact, I consider our direct marketing to consumers as a way of influencing the healthcare system. If you believe that prevention is ever going to amount to anything, you have to have far better information about the individual’s risk of disease. This is the key to limiting the burden of disease.”
The question is whether gene profiles are the future. In January 2008, a group of American geneticists warned, in an editorial in the New England Journal of Medicine, that there isn’t much clinical value in using them. The major illnesses that strike the general population – diabetes, cardiovascular disease, cancer – are complex. A series of genes is involved, but insufficiently understood. That’s true also of the multitude of environmental factors in play. It is uncertain whether gene profiles can provide a realistic picture of the risks that an individual faces.
“That’s pure nonsense!”
Now a tenor of anger has entered into his voice, and Stefánsson abruptly leans forward.
“A gene profile is as useful as any other form of screening. What would you do if it shows that you have a fifty percent chance of developing breast cancer? The statistics say that the disease is ninety-nine percent curable if it is diagnosed early, but is almost never cured if it is only discovered after it has spread. I think the percentage of the population that is cured would look far better if genetic profiles were used more diligently.”
The example gives me the cold sweats, but then I remember my fortunate gene variants and calm down. In the meantime, Stefánsson continues his soliloquy.
“We could also take cardiovascular disease. Our test has some markers for cardiovascular disease that provide a better prediction for the risk of heart attack than traditional measurements of cholesterol in the blood. For people who believe in preventative treatment, gene profiles are useful. But, of course, only if they react on the basis of the information.”
Even when he is speaking, he is restless in a way that is reminiscent of a hyperactive child. He constantly fiddles with his mobile phone and regularly demands more coffee – without at any point offering some to Farmer or to me. I hope to capture his attention with a point that I know will irritate him.
“We already know what we need to do to minimize our risk for the major prevalent diseases,” I say.
We should quit smoking, stuff ourselves with vegetables, make sure to exercise, and keep our weight down. But this knowledge does not necessarily make people change their lives, so why should it be any different if they know some numbers on their genome? For most people, chromosomes, markers, statistics – they’re all an abstraction.
Slowly, Stefánsson squints his eyes and considers me, the way you’d consider a colorful but disgusting insect.
“You talk about ‘people’ as though they were some sort of inferior race, and then you babble nonsense on top of it.” He taps his fingertips on the table to emphasize his words. “The experience from cholesterol measurements is that people take it quite seriously. About thirty percent of those who are found to have high cholesterol follow up by taking medication to lower it and changing their lifestyle. Another thirty percent react, but a little more halfheartedly.”
Of course. And forty percent ignore the information. In addition, cholesterol levels only say something about a single disease, whereas a gene profile provides risk estimates for several dozen different illnesses. Who can rationally take a position on that?
“You totally misunderstand. When you look at a gene profile, it is extremely rare to see an increased risk for more than one serious disease. I’m acquainted with one individual, whose profile I’ve seen, who has an increased risk of two major diseases. It is simply incorrect that you place an enormous burden on people by having them consider a heap of risk factors. And another thing is that when you have risk profiles for a significant part of a population you can as a society begin to shape targeted preventative health plans. This will probably be the only navigable way to limit costs in the future.”
I ask whether they should do gene profiles on all of us at birth. Stefánsson thinks that’s a very good idea. The more information there is, the greater the positive effect it can have for the individual over the course of his or her life.
“Listen,” he says, sounding tired of the discussion, “a lot of people wind up misunderstanding and misinterpreting this data. But since all life on this planet is based on DNA sequences and research is supplying an ever-deeper understanding of how these sequences affect the individual, it is inevitable that we will come to use it. It would be criminal to decide that people must not use this knowledge before we know exactly how it will affect them biologically, psychologically, and socially. You can consider what has been going on for the past few years as the clinical trial phase that takes place after a drug has come on the market. And, in this case, it is even driven by the consumers themselves.”
Now he apparently cannot sit still any longer, because out of the blue he asks whether Farmer has seen his latest series of pictures. Without waiting for an answer, he asks us to follow him. We trail him to a room down the corridor, which is utterly empty except for eight huge photostats leaning against the walls. They are extreme close-ups, taken on a beach not far from deCODEme’s offices, of pebbles and wet seaweed, blown up to a meter and a half and printed on aluminum plates.
Stefánsson describes the special lens they were taken with and talks about how much he loves to traipse about in nature with his camera as his only companion. I ask a little pointedly whether ordinary rockweed could be so brilliantly purple or whether his exposure has gone wrong. “That’s how it looks in reality,” he replies, lecturing me on the wondrous and unique Icelandic landscape.
“It’s the most beautiful place in the world.”
His pictures are quite good, and this irritates me for some reason. It doesn’t seem fair that you can be a model physical specimen, reach the top of the academic world, establish and run a high-tech business, and be a sensitive artist. In the back of my head, I once again hear my father’s speech about “good genes.”
How does some packaged, inactive genetic information specify that we become the people we are? Among these six billion base pairs, only a few percent separate a randomly chosen human being from an equally randomly chosen chimpanzee. What separates one human being from another is a mere half of one percent. How, I wonder, can these relatively few chemical changes have such wide-ranging significance and provide such different results as the kind, courteous, and curly-haired Farmer, the dauntless, sharp-tongued, black-clad Stefánsson, and me?
“Well,” he says, interrupting my train of thought. “You hav
e had a gene scan done yourself, I understand. Is there anything in it you want to ask about?”
BACK IN THE office, this time without a chaperon, I confide my test results and complain about my higher risk for arteriosclerosis. Since my first perusal the day before, I’ve read that the tested variant also increases my risk for lung cancer. It is up to thirty-three percent, which sounds quite overwhelming to my ears, but Stefánsson is not impressed.
“It doesn’t mean a thing, if you don’t smoke. Even though you are genetically predisposed, your risk falls to almost zero if you don’t smoke.”
I promise to stop but then say that I’m also nervous about my increased risk for basal cell carcinoma, the most common form of skin cancer. Here, too, there is no cause for special alarm.
“I’ve had one removed, it doesn’t mean all that much. You should just stay out of the sun.”
I already do but confess that there is something I don’t understand. I’ve been thin and bony my whole life, but now deCODEme reports that I have a higher than average risk of being overweight. Me, fat? I ask Stefánsson whether there could be a mistake in his test.
“No,” he replies with a conspicuous chill in his voice. “I myself have the same obesity variant as you but have weighed the same for many years – maybe because I exercise so much. Preferably, three hours a day. Otherwise I get depressed.” It may be the thought of the workout room, lined with dumbbells to beat off the blues, but he gets up and grabs a bottle of water from a small refrigerator in the corner of the office, and even offers one to me.
“Variants of multifactorial diseases will never give you a precise prediction, because they are complex and are not determined by a single gene. The outcome is powerfully influenced by environment. You have some gene variants but their penetrance; that is, to what degree they affect an individual, is a mystery. We can’t say what they do, whether some variants will be expressed or others will not.”