USA TODAY ARTICLE
The following article is from this mornings edition of USA TODAY:
Computers hasten search for cures
Genetic database may put drugs on shelves years faster
REDWOOD CITY, Calif. - Steven Evans-Freke and scientists at his company, Sugen, are trying to cure brain and ovarian cancers.
They've spent eight years and $100 million so far, a fraction of the 12 years and $300 million to $400 million or more that it traditionally takes to discover a drug and start making profits from it. Evans-Freke thinks his team eventually can shrink the drug-design process to seven years and, in the process, rake in profits from drug successes sooner and save millions of research dollars aimed at drugs doomed to fail.
To do so, Sugen and about 2,000 other biotech firms in this northern California biotech belt and elsewhere increasingly are supplementing the costly trial-and-error work of test tubes and tissue samples with high-speed computer analysis and genetic advances. "Any improvement in getting an innovative drug to market will have a huge payoff," says Elliot Sigal, vice president for applied genomics at Bristol-Myers Squibb.
That's where the new science of genomics and massive computer capability come in.
Pharmaceutical giants are forming partnerships and alliances with dozens of superspecialty biotech firms, each focusing on specific methods or disease processes. For instance, SmithKline Beecham has a large contract with Human Genome Sciences. Bristol-Myers Squibb works with Genome Therapeutics and Incyte Pharmaceuticals. Scientists are using computers to manipulate huge databases housing knowledge about every gene in the human body, to design potentially life-saving drugs at a speed unimaginable a few years ago. Some envision using this information to target drugs to patients most likely to benefit and least likely to suffer serious side effects.
Not everyone agrees that high-tech drug design will yield cures or treatments soon. In fact, despite billions of dollars poured into biotech firms in recent years, no drug developed using the latest computer techniques has yet reached consumers.
"It will take years to figure out how to apply all this technology," Sigal predicts.
But drug giants and biotech firms, many of them privately held start-ups, say the promise of such methods is compelling in the high-risk, high-reward pharmaceutical industry, where makers of the first drug of its kind on the market that treats a significant disease can expect a revenue stream of $300 million or more a year.
Sugen, a small, 7-year-old biotech firm that employs 200, focuses on one type of gene, the kind that controls communication from one cell to another and which Sugen scientists think plays an important role in cancers and other diseases, such as diabetes. When scientists at Sugen examine databases of genetic information, they look only at the role of those genes - leaving tens of thousands of other genes for other firms to study.
Specialization, speed
Such single-mindedness makes it possible to identify potential new drugs faster. Instead of first finding a substance in a plant or animal that could become a drug and searching for its use, companies identify a genetic target (such as the signal transduction genes), then search for a compound that affects it. It's a type of analysis that has only recently become possible.
Sugen has six cancer drugs under study, including one for brain cancer that is approaching the final steps before the company can seek government approval. A second drug, which cuts off the growth of blood vessels in cancer tumors, is also in early human trials. It aims at a biological target identified only five years ago. Two years later, clinical trials began. "It is extremely quick by pharmaceutical industry standards," says Evans-Freke, 46.
The human body contains about 100,000 genes. Each is responsible for a specific protein that, in turn, has a certain function in the body. Different proteins regulate disease processes. Once scientists discover a gene involved in a disease, they can begin to discover ways to prevent or treat the disease by manipulating the appropriate protein.
The genomics revolution was spurred in part by the federal government's massive Human Genome Project, aimed at mapping all the genes in the human body. In the early 1990s, former National Institutes of Health geneticist Craig Venter, founder of The Institute for Genomic Research, developed a technique of partially decoding genes using rapid computer analysis. Venter's work and that of others has led to the discovery and mapping of more than half of all human genes.
"We used to study one gene at a time. Now we can study the interactions of tens of thousands of genes" simultaneously, says Bristol-Myers' Sigal. Within hours, for example, scientists can take a sample of a prostate cancer tumor and figure out how its genes differ from healthy tissue, Evans-Freke says. With that understanding, researchers can search for or even design molecules that could block the disease process and become the foundation for drugs to treat or prevent a disease. "Five years ago, you physically couldn't have done that," he says.
More systematic
Randy Scott, president of Incyte, likens the genomics revolution to an auto mechanic who goes from understanding 2% of a car to 80%.
The old way, drugs were discovered somewhat by chance. Scientists would identify part of the process that caused a disease, then test thousands of compounds against it to see if any of them changed the process. In many cases, even the scientists who created drugs on the market today don't understand why they work.
The ability to catalog, analyze and deliver all the genetic information available is fundamental to the new model of drug creation, Scott says.
"Right now, we're trying to cure cancer, but we don't understand what makes a cancer cell a cancer cell," he says. But with genomics, "probably in 10 years we'll understand the molecular basis of every major human disease."
Scott says Incyte's main product is no longer pharmaceuticals but genetic information, which customers can access for drug research.
At Seattle-based Immunex, just such a database search led scientists to a protein they call TRAIL, which appears to play an important role in a wide range of cancers. The drug they are studying as a result could be an effective new cancer fighter and among the first produced out of high-tech computer processes.
Immunex, started in 1981, has 850 employees and 1997 revenue of $185 million. Like most research-heavy biotech firms, the company hasn't yet shown a profit. But promising drugs in its pipeline give company leaders and investors in Immunex reason for optimism.
Doug Williams, senior vice president at Immunex, says it took only months to identify TRAIL using gene-based technology. Traditional methods probably would have taken four to five years. Not only does this save millions of dollars a year in research costs, it gives Immunex proprietary rights to the gene. If another company wants to develop a product based on TRAIL, it must license the information from Immunex.
Soon, Williams expects to begin human trials of TRAIL to determine its effectiveness in treating cancer in humans. If all goes well, Williams estimates Immunex will bring the new drug to consumers two years earlier than previously possible. That two years can mean an additional $600 million or more in revenue.
Designing for speed
Kwang-I Yu, CEO of Paracel, a Pasadena, Calif., firm that designs computers for biopharmaceutical analysis, sees the world of new drugs as "low- hanging fruit" just barely out of reach. He's optimistic that computer analysis can replace or at least limit the need to perform safety tests of new drugs on animals, a step now required before the Food and Drug Administration allows human studies.
Kwang-I's company is speeding up computers by limiting their abilities. Unlike a desktop PC, Paracel's computers can analyze DNA but cannot do word processing or a variety of functions. Paracel's latest computer has up to 7,000 microprocessors that can do 10 billion genetic calculations in one second, vs. a few million for an ordinary computer workstation, Kwang-I says.
Ultimately, the genomics-driven biopharmaceutical industry could result in new, highly specific drugs capable of treating diseases in ways never before possible.
A study by Andersen Consulting finds that most pharmaceutical companies expect their drug discovery timelines to be halved within the next few years. But until that happens and drugs resulting from those technological advances start reaching consumers, skeptics will remain.
"It's hard to imagine how understanding diseases at the molecular level at some point will not lead to dramatic advances in therapy," says Charles Engelberg, biotech analyst at AmeriCal Securities. "But so far, it's all sort of theoretical."
Evans-Freke and his counterparts are undaunted. "If you reach the finish line with a real product on the market, the returns for everybody - including the patients - are very considerable."
By Doug Levy, USA TODAY |
