Advances in genomics, molecular biology, and bioinformatics are converging to form an environment rich for pharmacogenomic discoveries. But who will till the soil?

Twenty years ago this month, record industry manager Ken Kragen and at least 5 million participants pulled off "Hands Across America," a fund-raiser for homeless charities that featured people holding hands in a (mostly) unbroken chain from New York to California. The few reported breaks in the chain didn't end up mattering very much when considering the scope of the undertaking and the millions of dollars raised.

Pharmacogenomics—an emerging science that combines the study of the human genome with drug discovery efforts—could benefit from a smaller-scale but nonetheless

click the image to enlarge   Allele-Specific Primer Extension (Source: Richard Weinshilboum, MD)

impressive coordination among the academic and pharmaceutical research communities. To listen to pharmacogenomics pioneer Richard Weinshilboum, MD, professor of pharmacology at the Mayo Clinic, Rochester, Minn., is to believe that the breaks in the pharmacogenetics chain are bridgeable, with knowledge that may already be within researchers' grasp, if only enough experts could join together in the search for solutions.

To that end, Weinshilboum and others at academic institutions and the National Institutes of Health have come together to form the Pharmacogenetics Research Network (PGRN) to collect and analyze the best of the best data. "How are we going to move forward to find ways to bring this exciting new science to the bedside and move it toward keeping the promise of tailoring therapy to the individual patient?" Weinshilboum asks. "First, we're going to need collaborative interactions of a variety of disciplines: basic genomic science, statistical genetics, bioinformatics, and translational clinical applications, bringing groups with different and complementary expertise together."

Such information-sharing among academics has fueled a wide range of pharmacogenomics successes in recent years, largely in areas such as identifying single-nucleo-tide polymorphisms (SNPs), small genetic variations within a person's DNA sequence that explain why certain drugs metabolize differently in some patients. Re-searchers have identified polymorphisms that occur in key pathways, such as in cytochrome P450 (CYP450), the set of enzymes responsible for metabolizing many drugs in the liver, and in thiopurine methyl transferase (TPMT). On a small scale, pharmaceutical products such as Herceptin (trastuzumab, used to treat breast cancer) and Gleevec (imatinib, used to treat tumors) have been developed that target individuals who over-express certain enzymes.

The Immediate Future of Academic-Pharmaceutical Partnerships The more researchers learn about how people with different genetic structures metabolize drugs, the more it would seem pharmaceutical companies could dig through the thousands of compounds that have been rejected due to toxicity issues and find some that may work for people with genetic variations. "I would’ve thought that there would be a strong impulse to try to rescue compounds that have failed during clinical trials as a result of genetic variations in adverse drug responses,” says Richard Weinshilboum, MD, professor of pharmacology at the Mayo Clinic. "But this isn’t very compelling to the pharmaceutical companies. They feel that because of the ticking of the patent clock and the amount of effort that would be involved in rescuing a compound, they’re just not prepared to do that. This is particularly true of big pharma, which is still to a large extent focused on the blockbuster mentality" But Weinshilboum believes there is a potential avenue of academic-pharmaceutical partnerships that we might see very quickly: postmarketing safety surveillance. "This would involve the academic community working closely with the pharmaceutical industry once drugs are already broadly available to protect against the kind of terribly unfortunate situation that occurred with Vioxx,” he says. "The proprietary nature of the information the companies have continues to be a barrier, but I think we’ll see more opportunities for this kind of partnership in the future."

"When you look at what has actually been validated clinically from pharmacogenomic research, in most cases you're talking about the cytochrome P450 genes and others for which clinical utility has really been established," says Elaine Weidenhammer, PhD, associate director of business management for Nanogen Inc., San Diego. Nanogen offers a series of arrays and reagents for genotyping CYP450 and other drug-metabolizing enzymes, including the DrugMEt pharmacogenetic test offered through a partnership with the Finnish company Jurilab. DrugMEt provides simultaneous detection of up to 29 SNPs of eight enzymes that are critical to the metabolic pathways of many prescribed drugs.

"Pharmacogenomics will help us understand a person's susceptibility to disease and then combine that with an understanding of how they're likely to respond to drugs," says Weidenhammer. "Ultimately, these will be used prognostically in combination with the genetics and the protein biomarkers. Then you can really get into helping patients manage their own healthcare prior to even having a disease phenotype developed."

DrugMEt is a glass slide microarray geared toward the research labs of potential pharmaceutical partners who want to incorporate pharmacogenomics testing into drug development and clinical trials. But therein lies the problem. "To date, we have not had significant interactions with pharmaceutical companies," Weidenhammer says. "But we certainly are anticipating that we will, especially with the markers that Jurilab has identified in their founder population. We anticipate that there will be some novel drug targets arising from those studies."

Waiting on big pharma
If Herceptin, Gleevec, and Erbitux (for treatment of colorectal cancer) signified the first wave of individualized medicine, the second wave has been painfully slow in developing. "Since the concept of pharmacogenetics has been around since the middle of the 20th century, why isn't it having a greater impact in the clinical arena?" asks Weinshilboum. "There are a variety of answers. It's taken a while for the science to mature, but it has accelerated significantly as a result of the completion of the Human Genome Project. Now, we have the tools to move everything forward at a much faster pace by putting the advances in genomic science together with the concepts of pharmacogenetics and advances in molecular pharmacology."

click the image to enlarge   Representation of pharmacogenomic “players” and their relationships. (Source: Richard Weinshilboum, MD)>

For example, advances in microarray technology and in assay development have dramatically improved throughput, performance, and cost. Bead-based formats used in products such as BeadArray from Illumina Inc., San Diego, Calif., provide the accuracy of multiple measurements and statistical power even when using a small number of arrays. Such high-throughput research tools enable researchers to sift through tens of thousands of genes and hundreds of thousands of variations in the human genome.

The problem thus far has been the lack of any systematic involvement from the major pharmaceutical companies. "The major initiatives for the development of new drugs come typically from the pharmaceutical industry," Weinshilboum says. "Up until recently, the pharmaceutical industry had very little in the way of incentives to include [pharmacogenomics] in what they did. That's not to say that there isn't an understanding in the industry of individual variations in the response, which range from lack of efficacy to adverse drug responses. But the only thing this offered the pharmaceutical industry in the past was market segmentation. That's not a very good incentive from a financial point of view."

Without investment and follow-up from pharmaceutical manufacturers, current pharmacogenomics research seems like a baseball pitcher throwing warm-up tosses to the catcher in perpetuity, with no batter in sight.

Weinshilboum and colleague Liewei Wang, MD, PhD, assistant professor of molecular pharmacology and experimental therapeutics, Mayo Clinic, recently wrote a review article on the history of pharmacogenomic breakthroughs. "When we went back over the series of the major advances, beginning with the early work, we were forced to say that all the major breakthroughs we teach our students about have come from the academic side," Weinshilboum says.

Now, an increased focus on pharmacogenomics advances by the US Food and Drug Administration (FDA) may change that. In addition to concern about how to turn a profit on a drug that works for only a small subset of people, pharmaceutical companies have been troubled by how to validate pharmacogenomics data submitted to the FDA as part of a drug approval process. But guidelines released last March specified that voluntary submissions of pharmacogenomics data would not be used for regulatory decision-making.

"The vigorous involvement of the FDA has gotten the attention of the pharma industry in a way that academic centers are not in a position to do," says Weinshilboum. "But whether or not it will be enough to incentivize the pharmaceutical industry is still an open question."

Looking beyond SNPs
Far from simply waiting on the sidelines until pharmaceutical companies get into the game, academic researchers keep plugging away at the elusive connection between genetic variation and response to drugs (see "Recent Advances in Pharmacogenomics Research," below).

While most pharmacogenomics research has focused on the identification and study of SNPs, pockets of researchers have begun to widen the search, including Charles Lee, PhD, director of cytogenetics at the Dana-Farber Cancer Institute and assistant professor of pathology at Brigham and Women's Hospital, Boston. Lee and colleagues have

Recent Advances in Pharmacogenomics Research Academic researchers are pushing advances on a startling variety of pharmacogenomic fronts. Here is a brief sample of some avenues of pharmacogenomics research: • A recent research report in the New England Journal of Medicine identified a single genetic mutation that accounts for more than 20% of all cases of Parkinson’s disease in Arabs, North Africans, and Jews. This is a big development in an unlikely place, as most experts believed the low-hanging fruit of pharmacogenomics research would be in leukemia, epilepsy, heart conditions, or cancer. • Researchers at the Mayo Clinic have used pharmacogenomics to develop a test and treatment for an inherited kidney disorder (type I primary hyperoxaluria) that can cause organ failure in children and young adults. The researchers found that a genetic mutation allows vitamin B6 to benefit patients by preventing kidney stones. They are now using the finding to develop a genetic test to predict which patients are best suited for this treatment. • Mayo Clinic researchers are also behind a new test that identifies patients at high risk of adverse reactions to the colon cancer drug Camptosar (irinotecan HCl). The test detects a change in the DNA of a gene that encodes for a protein affecting metabolic response to irinotecan, which has tremendous benefits in fighting tumors but also has tremendous—sometimes lethal—toxicity for many patients. The new test gives scientists advance knowledge of the individual risk for irinotecan toxicity. • Japanese researchers are working on further defining the link between the CYP2A6 enzyme and nicotine addiction and withdrawal symptoms in smoking cessation. The next phase of that research is to individualize smoking cessation programs based on CYP2A6 genotypes.

found that copy number variations (CNVs) frequently overlap with genes and could explain why people respond differently to drug compounds.

"We came upon CNVs by accident rather than as a result of a dissatisfaction of SNP research in pharmacogenomics," says Lee. Serendipitous or not, research using CNVs has captured the attention of some of the nation's leading researchers and spawned work from renowned researchers Michael Wigler, PhD, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., and Evan Eichler, PhD, University of Washington, Seattle, among others. Eichler's work on CNVs by sex and the recent findings of genetic mutations in different races that contribute to Parkinson's disease are not only fueling future genomics research—they're likely to lead to a greater understanding of human evolution.

Next up for Lee and his colleagues is an in-depth comparison between what we know about CNVs in relation to what we've learned about SNPs. "We'll be looking at a comprehensive cataloging of CNVs in the human genome, starting with the HapMap samples," Lee says. "That will also allow us to look at SNPs and CNVs in the same ‘normal' individuals."

Next target: the "ego" gene
Looking to the future of pharmacogenomics research, Weinshilboum first takes a moment to marvel at the past. "The public doesn't appreciate that in addition to a genomic revolution, we've had a therapeutic revolution," he says. "When I was in medical school and I saw a kid with acute lymphoblastic leukemia—the number one cancer of kids—those children were uniformly dead within one year. Today, we cure 85% of those kids. That's purely with drug therapy."

The bar is set even higher with pharmacogenomics, because of the legitimate possibility of truly personalized healthcare. "My guess is that the long-term implications will involve selection of those patients who are going to have the desired therapeutic response," Weinshilboum says. "Nothing in medicine is ever going to be absolute, and certainly pharmacogenomics isn't, but it will help us better tailor therapy on the front end."

When asked if he could have one wish that would further the development of pharmacogenomics faster than anything else, Weinshilboum doesn't hesitate. "I wish we could knock out the human ego gene," he says. "The technology is wonderful, but we need to find ways to bring together the multidisciplinary teams with expertise that complements each other. This is truly a multidisciplinary, multidimensional problem that will require the gifted academic and scholarly clinicians working together with statistical geneticists and genomic experts and those who understand the molecular pharmacology. Putting those teams together, overcoming the individual barriers—and institutions have ego structures, too—to bring together the teams we need . . . that's the major challenge."

Nanogen's Weidenhammer agrees. "It's always this struggle with that gap that exists between discovering a marker and assessing its potential correlation with response or disease development in 20 patients, and then really establishing clinical utility for that marker," she says. "That is a problem both for diagnostics and for therapeutics. We don't yet know quite how to handle that. We hope that the collaborative programs trying to bridge this gap will help us get to the next step."

About the Author
Bill Schu is a freelance writer based in Bloomfield, N.J.

This article was published in G & P magazine: Vol. 6, No. 3, April, 2006, pp. 16-19.