Great article on Genomics! -->
Genomics Gets Down to Business
by Gwen Kinkead June 2001, Worth magazine
The mapping of the genome? Last year's miracle. New and old players in the drug industry are now staking out positions in the science of genotyping, which is on its way to giving us a powerful Understanding of our genetic destiny and changing the face of medicine.
In the race to map the human genome, government-funded researchers competed furiously with private companies to draw a picture of our common genetic heritage. Now a new marathon, mapping the body's collection of single-nucleotide polymorphisms — the variations in the genetic code that make my biology different from yours — is about to be completed. This time, business has made sure to keep the work, and hence the potential profits, away from the government. That's because single-nucleotide polymorphisms, or SNPs (pronounced "snips"), are the gateway to personalized medicine — drugs designed for each person's individual genetic makeup. The commercial exploitation of SNPs represents a colossal new opportunity for the $330 billion worldwide drug industry.
This month, the industry-funded SNP Consortium will wrap up its two-year, $30 million project to map the body's SNPs, and the nascent science called genotyping will accelerate in its rush forward. The map will provide a readout of 3 million of the many millions of SNPs in the genome. It is a tool that will be used to identify those points in the genome that underlie drug response and susceptibility to disease. That's the essence of genotyping, which is itself the most provocative part of the science of genomics. Think of a genotype as a kind of genetic horoscope. One day, we may all carry our own genetic profiles in our pockets, on medical-history cards or DVDs.
Within three to five years, drug companies expect, doctors will be able to read a person's SNPs before prescribing a drug and, thereby, learn whether the drug will be safe and effective or harmful. Genotyping will also reveal the SNPs that indicate susceptibility to major diseases. Those are two big steps down the road to personalized medicine. Pharmaceutical companies are crazy to know our SNPs because that knowledge has the potential to make each of us a separate market. So long, one-size-fits-all drugs and hit-or-miss medicine.
With promises such as these, it's no wonder so many companies are piling into the analysis of single-nucleotide polymorphisms. They include the pharmaceutical companies, of course, both new and old, and also a small explosion of specialist firms whose work could scarcely have been imagined before the SNP project got under way two years ago: genotypers, which perform the technical work of analyzing people's genomes via their DNA; toolmakers, which create the tests to explore the microscopic inner workings of genes that genotypers use; information specialists, which are building the supercomputers and writing the algorithms and software needed to mine the mountains of data on human biology created by the genotypers; and DNA harvesters, the organizations and companies that ask people around the globe to offer their DNA for study.
They're all focused on less than 0.1 percent of each person's genome, the SNPs. That's the essence of anyone's uniqueness; other than our SNPs, we're essentially identical. Every person has roughly 30,000 genes but millions of SNPs. Of these millions, mere thousands are medically relevant: They either precipitate or increase our risk for a disease, or they are markers for a mutation. A SNP that has been shown to indicate susceptibility or predisposition is referred to as a SNP assay.
In many scientific circles, genotyping inspires an almost giddy optimism. Francis Collins, director of the National Human Genome Research Institute, which funded the government-sponsored team that raced Celera Genomics to crack the genetic code, describes genotyping as history waiting to happen. Collins predicts that by 2010 we'll take genetic tests to determine our risks for a dozen complex hereditary diseases and that by 2020 the first designer medicines will be administered. By 2040, Collins says, "individualized preventive medicine will be routine."
That's the kind of thinking that attracts money — it has, for example, made Bill Gates the biggest biotech investor in the world. But not everyone is sanguine. Some people in the industry worry that consumers may not embrace something that requires them to give up their genetic secrets to strangers. They say that the distance between what we know now and any kind of personalized medicine is enormous. And they raise questions about the ethics, the cost, and even the feasibility of SNP analysis.
Business and science are arm in arm; as they advance together, they're sometimes likely to stumble together too. Here's where they stand in the early moments of a long and amazing journey.
Big Pharma
Even the best drugs don't work for everyone. They perform as expected for most patients but have no benefits for others and have adverse effects on at least some. That's why the first SNP-based order of business for the drug companies is to identify the SNPs that govern response to drugs. Doing that, the companies believe, will double the number of drugs that get approved, which matters a great deal when the cost of developing a drug and seeing it through the approval process now averages $500 million. The drug companies will certainly start digging for the SNPs implicated in common diseases, including cancer, but that will take time. In the meantime, they think, they'll hit an economic jackpot in the area of response.
A small but crucial part of anticipating drug response is avoiding the forced withdrawal of drugs that have already made it to market. Allan Roses, head of genetic research at drugmaker GlaxoSmithKline (NYSE: GSK), points to the painful example of Lotronex, a Glaxo drug that was yanked from the market last year. It was withdrawn after 43 people suffered side effects, sometimes requiring resectioning of parts of their bowels. Another five people died. "A couple of hundred thousand women over a four- to five-month period took this for irritable bowel syndrome," Roses says. "If we could identify, using pharmacogenomic genotyping, whether a woman fits genotypically into the 43 or into the couple of hundred thousand, we could just tell the people with the genotype of the 43, 'Don't take the drug.'"
Glaxo is working on the first SNP drug-response test, for abacavir, an HIV drug sold under the trade name Ziagen. Abacavir causes a serious, sometimes allergic reaction in about 5 percent of patients who take it. "We have collected blood samples from people who have had the hypersensitivity reaction," Roses says. "The blood from these volunteers, who have signed informed-consent documents, will be used to develop a test for adverse events." Glaxo will call on a genotyping company to sequence the DNA in the blood of these individuals to find the culprit SNP or groups of SNPs. "Then, we and the FDA agree that before new patients are given abacavir, their doctor tests them to make sure they are not the genetic type to have side effects. The doctor takes a blood spot from each new patient and sends it to a genotyping lab for overnight testing. If the patient's SNPs match the SNPs that cause an adverse response, Glaxo notifies him or her the next day not to take the prescription.
"If it works, we'll do it on all our drugs," Rose continues. "We will store blood spots with trusted third parties. Everyone in the industry is waiting for our abacavir test: We will be proof of principle. Finding the people and taking their DNA samples cost us millions, but this is a win-win: We don't lose $500 million when someone has an adverse reaction, patients get safer drugs, and the regulators like the FDA have objective criteria by which to exclude patients and so ensure the safety of the population. The FDA will go from scratching its head to requiring it in eight to 10 years."
The Food and Drug Administration is mum about whether it will allow genotyping tests for drug response, but most of the industry is prepared to crash-start genotyping this year. The savings could come to $33 million per drug by 2010, according to Front Line Strategic Management Consulting, a Silicon Valley biotech consultant, because adverse events will drop and clinical trials will be limited to people who have been hand-picked by genotyping to respond well to a new drug.
Analyzing people's genomes will also lead to more diagnostics. "For instance, it's likely we will be able to stratify adult-onset diabetes into five or six different types of disease," says Lee Babiss, vice president of preclinical research and development at the pharmaceutical giant Hoffman-La Roche. "We will be able to test your genes and see which you have."
From diagnostics will come new therapies, says Ken Conway, president of the predictive medicine subsidiary of Millennium Pharmaceuticals (Nasdaq: MLNM). Millennium, the biotechnology star in Cambridge, Massachusetts, wants to be the leader in personalized medicine. It has a market cap of $7 billion because it is considered the smartest biotech at finding new drug compounds, although it has yet to sell a drug or make a profit.
SNP-based medicine, Conway predicts, will transform the treatment of cancer. "A simple blood test will diagnose ovarian cancer, for example, before symptoms," he says. Will people believe they are sick if a test says they are but they don't feel sick? "They will believe once their ovaries are removed and they are cancerous," Conway says. "With another test, we will match the right medicine to your genotype and, with a third test, determine if you are having a molecular response to that medicine. If we see molecular changes before you start to feel better, you are getting better. If not, we will find another drug, eliminating weeks of waiting for lab results."
Genotypers
There's no such thing as a well-established genotyping company. A lot rides on the ingenuity of little firms like Genaissance Pharmaceuticals (Nasdaq: GNSC), in New Haven, Connecticut. Genaissance, still in its first year of public trading, looks for SNPs that predict drug response. It is testing the DNA of 600 patients in an effort to unravel why they respond differently to four top-selling cholesterol drugs.
Orchid BioSciences (Nasdaq: ORCH), of Princeton, New Jersey, has also bet the farm on SNPs. It plans to grow with the market — first finding SNPs in DNA sent by Big Pharma to evaluate how people metabolize medications, then selling SNP diagnostic kits, and then using SNPs to come up with new drugs to license. Orchid has 20,000 SNP assays in its proprietary databases. "SNPs will be studied for 50 years," says Mike Boyce-Jacino, the company's vice president of technology.
Celera Genomics (NYSE: CRA) is the giant out front, the celebrity company that last year helped to sequence the genome. Its strategy is slightly different. Instead of sequencing SNPs for drug companies doing drug-response and diagnostic testing, it is conducting basic research to uncover clinically valuable SNPs, which it will ultimately sell to drug companies for use as components in other genotypers' tests. It has the largest SNP database around, with 4 million to 5 million SNPs, though the number itself isn't terribly meaningful given the small percentage that will prove to be medically significant. "We're looking, like the Marines, for a handful of good SNPs," says Celera's contentious founder, J. Craig Venter. "Some will potentially be worth hundreds of millions of dollars, the ones that can be turned into diagnostics to assess a person's risk for common diseases or for a toxic reaction to medication."
Celera is focusing first on three disease categories: cancer, inflammatory diseases, and diabetes. Then it will systematically grind through all common diseases, peering into DNA from thousands of donors, seeking SNP assays to sell. Celera also has a mammoth crash research program in proteomics, the study of protein-to-protein interactions. (Proteins do the work of the body, carrying information from cell to cell and causing the body to perform essential functions. Proteomics is complementary to SNP analysis and is the next level of inquiry in genomics.)
Celera, Orchid, Genaissance, and their competitors, a crowd of about 25 small companies, use different tools to identify SNPs in DNA. The methods include mass spectroscopy, biochips (also called micro-arrays), and others. Some have proprietary technology; others buy it from the toolmakers.
Toolmakers
The biochip is genotyping made manifest. The size of a dime, a biochip looks like a microchip, but instead of electronic circuitry, it's coated with single-stranded DNA. A single strand of DNA is sticky; it wants to associate with its sister strand, or counterpart, to form a double helix. Its nucleotide bases always bind in the same order — adenine with thymidine, guanine with cytosine — across the helix. When there's a mismatch, that's a SNP (see "SNP Basics")
Thousands of strands of a reference DNA are built up on the chip with photolithography or ink-jet spray, two common methods of application borrowed from the semiconductor industry, or by pin spotting, by which a robotic steel pin dips into liquid DNA and drops a minute amount on a glass slide. This is repeated over and over until a thick hedge of strands bristles on the slide.
The DNA to be tested is extracted from blood or tissue and single strands are removed. Each strand resembles a clear glob of mucus before it is mixed with fluorescent dyes and pipetted into the chip through a tiny hole. It seeks out its complementary nucleotide base pairs in the DNA on the chip to form a double helix. Even a one-base difference in the nucleotides of the sample DNA is enough to make the fluorescence glow slightly less in one spot when a computer scans the chip, confirming that the sample contains a certain SNP. In this fashion, one chip can read thousands of strands simultaneously. For instance, a chip might test whether you have the SNPs that indicate that you will develop toxicity in the liver if you take the class of diabetes medication called glitazones. Eventually, chips could test hundreds of SNPs to see what diseases you harbor.
The FDA is just now catching up to this field. Regulators have toured the industry leader, Affymetrix (Nasdaq: AFFX) of Santa Clara, California, to observe the manufacturing process. The FDA hopes to determine, within the next few years, how it will regulate this industry on which genotyping and genomics depend.
Genotypers and their tools — the various kinds of chips, the microscopic beads developed by Lynx Therapeutics (Nasdaq: LYNX ), and other technologies — are the bottleneck in genotyping. None has yet demonstrated low enough costs for automated mass screenings of genotypes. The current cost, from test tube to printout, is between $1 and $2 per genotype. Industry analysts agree that for genotyping truly to take hold, the cost has to drop to pennies or tenths of pennies. Orchid wants to get there with the "arrays of arrays" it will introduce in 2002. These are plates of 384 wells, each of which functions as a biochip. Within each well, tens of thousands of SNPs can be analyzed at once with robotics — the sort of supercharged mass screening the industry wants.
Affymetrix, the pioneer of biochips and the gold standard for accuracy of testing, has the most powerful but not the most versatile single chip. Its genotyping chip currently analyzes 1,500 SNP assays, but founder Steve Fodor says that will change. "We'll genotype 50 people," he says, "and put SNPs we find that are implicated in common diseases on chips and sell them to drug companies for diagnostic tests. We have a new version of this technology not yet ready for release, with 20 million to 60 million pieces of DNA on it. So we can imagine entire chromosomes on one chip. It would take a couple of hundred of these to study the entire genome."
Where is all this going? "The endgame for all these toolmakers is to have their SNP chips in the handheld devices that doctors will use to diagnose and treat you in five to seven years," says Darren Mac, a biotechnology analyst with Gruntal & Company. These little boxes will test your DNA with a disposable biochip cartridge containing panels of SNPs. Once your blood is put inside the chip, the device will check your SNPs against the panels, and any matches will be visible through a screen or a window on the device, indicating, for instance, which asthma medication will help you most or that you have hepatitis C. The leader in this field is Clinical Micro Sensors, a division of Motorola (NYSE: MOT). That company's first-generation product, the eSensor, is scheduled to ship to researchers this month. It looks something like an overcomplicated PDA — except that a PDA can't yet tell you what might eventually kill you or how that might be prevented.
IT Solutions
Organizing the gusher of data from scans of individual genomes will create another enormous market. IBM (NYSE: IBM), for example, is jumping headfirst with a $300 million investment in a new-generation supercomputer, the world's largest, to manage, store, and interpret the flood of information, and in developing new software and partnerships with genotypers. Medical records alone will be huge: Each person's genotype represents an overwhelming amount of electronic information. "The raw data are growing superexponentially," says Caroline Kovac, head of IBM's new life-sciences division. "It's exactly true that we have almost too much data, more than people know how to digest and extract understanding from. And unless you can do that, you'll never be able to put it to work in decision making; you'll never be able to get value out of it. So that's where information technology is absolutely critical: It's what's going to allow you to get that high-end analysis." IBM's new $l00 million supercomputer, called Blue Gene, will perform complicated analyses of protein structures and gene pathways associated with disease, all necessary steps to arrive at personalized medicine.
DNA Harvesters
Where is all this DNA from? Not this country. Thus far, most Americans are too spooked about genetics to donate. Following the news last summer that the human genetic code was mapped, a Time magazine poll found that 75 percent of the 1,218 Americans surveyed did not want insurance companies to learn the mysteries of their DNA, and 84 percent wanted that information withheld from the government. People fear that their genetic predispositions to conditions such as mental illness could be used against them. Let's say you gave DNA to a drug company to genotype and it discovered that you may develop a life-threatening disease. Would you want to know? Is it obligated to tell you?
DNA for genotyping is largely bought from monster DNA data banks overseas: deCode genetics (Nasdaq: DCGN), of Reykjavik, Iceland, is the first company to turn an entire country into a data bank. It has signed nearly all the 275,000 inhabitants of Iceland into a DNA database. Iceland, which has a small, relatively homogeneous and relatively isolated population, is a happy hunting ground for deCode's geneticists to trace the inheritance of disease traits for generations, thanks to the country's 100 years of socialist medical records and 1,000 years of oral storytelling outlining family trees. Estonia is another country that has turned into a gene bank. The Wellcome Trust is trying to enroll a half million volunteers for a gene bank in England.
Only one U.S. company, Genomics Collaborative, of Cambridge, Massachusetts, has a sizable DNA vault. The company sweeps in DNA from doctors, scientists, and hospitals. "We are signing up 5,000 to 7,500 new subjects a month, about half from the U.S.," says Michael Pellini, president of the firm. "We have 80,000 samples that we sell to researchers and also study ourselves. We don't tell people how we will use their DNA; we ask permission to do various procedures. Other people in this field are having trouble because they just jumped in, expecting recruitment would be easy. We spent more than two years building our network of physicians."
DNA Sciences, a glamour-puss from Fremont, California, has signed up about 5,000 volunteers since opening last August. It registers people via its Web site, DNA.com, which makes this pitch: "Ordinary people doing extraordinary things.... It's nothing less than a chance to become part of history."
First Genetic Trust, of Chicago, is an example of the DNA vaults that are springing up around the U.S. to store genetic information. At the moment, it stores the genetic information drawn for a clinical trial at Memorial Sloan-Kettering Cancer Center. Down the road, in perhaps five years, it will do something extraordinary: give each of us a chance to lock up a digitized readout of our DNA. First Genetic Trust will guard the genotype so that it is available for research when, but only when, you give the okay. "We are at the dawn of a new era," says founder Arthur Holden, formerly head of the SNP Consortium. "There's a need for legislation here," Holden says, "but also a role for private companies in order to give people peace of mind and control so that their genetic information will be used by pharmaceutical companies only in the manner they consent to. Our business allows research to go forward and privacy issues to be addressed."
There is little space in the drug industry between science and business, and in genotyping that is especially true. All these companies are dependent on science that is emerging and evolving daily. And they're dependent on one another. It doesn't make it easy to identify leaders.
The picture is complicated further by the ethical, cultural, and political pressures that will inevitably come to bear. Americans might prove very stubborn about their DNA. "Most of the drug companies and biological visionaries in genotyping are concentrating on the technology — can we get genotyping done and cheaply enough — but the most important issues here revolve around genetic privacy," notes Jeff Handen, a biotech consultant at PricewaterhouseCoopers. "The industry knows there is a problem. They're very worried about the social impact of this new technology and whether it's going to bite them in the rear."
Nor is everyone convinced that genotyping will produce the results its proponents describe. Richard Horton, editor of the medical journal The Lancet, says genomics will never find a direct cause of most common diseases because it doesn't take into account the role of the environment in influencing disease. "That message is not one that many scientists want the public to hear," he says.
Perhaps the road ahead is bumpy or worse. But that isn't going to stop medicine from looking longingly at the raw material written in DNA. Knowledge of its precious secrets is going to shape us somehow in the future.
And it will shape the economy, perhaps as forcefully as the computer business has for the past decade. Which leads us back to Bill Gates, who owns shares and interests in a breadth of biotech companies, most of them small. His biggest stake is in Icos (Nasdaq: ICOS), a biopharmaceutical firm with three products in Phase III testing.
A less risky way to bet that great research turns into revenue is to follow the toolmakers and IT specialists. They will thrive no matter what, because they will be needed for the next human-genome marathons, such as the studies of proteomics. The story has just begun to be told. |
