BACK TO THE FEATURE INDEX

The aftermath of the human genome project promises gene-based medicines, with newly discovered protein targets beckoning to R&D managers and prompting expanded research budgets. A catalog of common polymorphisms, many think, could open the door to so-called "personalized medicine." Patients that currently cycle from one antidepressant drug to another in search of relief, or from cancer drug to cancer drug in a race for survival, may find assistance from researchers who are working hard to trace links between drug responses and genetic variation. In technical circles, this area of research is known as pharmacogenomics.

Most clinical trials now include a pharmacogenetic component, or at least a collection of blood samples that can be used for future analysis. Early work focuses primarily on the drug-metabolizing cytochrome p450 enzymes, some variations of which can significantly alter patients' response to drug regimens.

Despite its scientific and medical potential, the field is being embraced with some reluctance. Companies worry that genetically targeted therapies in effect reduce the market for a given drug. After all, if a test shows that 60% of patients respond strongly to an antidepressant and the rest respond poorly or not at all, that extra research investment has reduced the size of the drug's market by 40%. It might be good science, but it isn't good business.

The end of blockbuster drugs?

Will pharmacogenomics spell the end of the billion-dollar blockbuster drugs? No, says Michael Murphy, president and CEO of Gentris Corp. of Morrisville, North Carolina. "A blockbuster drug by definition works in the majority of the patient population. If it has significant adverse events, the drug probably won't generate blockbuster status anyway."

Even if companies remain hesitant, other forces will push pharmacogenomics, believes Andrew Kim, an analyst at FrontLine Strategic Consulting Inc., now a part of Navigant Consulting Inc. Kim is the author of a 2002 pharmacogenomics industry report. "This is something that will be taken up by patient advocates."

In fact, pharmacogenomics is a potential salve for more than just patients. Companies see potential for this new science to salvage at least a few of the many drugs that fall through the cracks in the development pipeline. Many candidates never make it to Food and Drug Administration (FDA) review because of serious side effects or efficacy in only a limited number of patients. If these problems could be tied to genetic variations, these drug candidates could be given new life as therapies for patients with particular genetic profiles.

Pharmacogenomics is also being used to streamline the drug development process and reduce the cost of clinical trials. Genetic tests that reveal toxicity and side effects early on could help eliminate red-herring molecules before they eat up too much of the R&D budget. When it comes time for clinical trials, prescreening of patients most likely to respond could reduce the number of patients required for a clinical trial and produce more robust results and a swifter path to approval.

ChondroGene Ltd. of Toronto, Canada, is banking on just that. The company is developing diagnostic tests for early stage osteoarthritis (OA) and genetic markers that indicate disease severity and future risks, as well as the most appropriate treatments. Clinical trials in OA are particularly costly because the condition is difficult to detect in its early stages; clinical trials require long terms and large patient populations. "If we can quantify the severity of the disease [using ChondroGene's genetic tests], companies won't have to run clinical trials quite as long, so it will save them a lot of time and money," says Bruno Maruzzo, who is the company's director of corporate development.

High demand for talent

As you might expect, all of these research problems spell jobs. Even in a poor biotech job market--2003 was one of the worst years ever for employment in the biotech industry--pharmacogenomics trainees are finding that opportunities abound. "There isn't a pharmaceutical company that doesn't want these people," says Dave Flockhart, professor of medicine at the Indiana University School of Medicine, who directs a clinical pharmacology fellowship program for M.D.s. Fellowship participants "are getting calls from headhunters 6 months or a year before they've finished" the program, he says.

The trouble is, there aren't nearly enough training programs to satisfy the demand. "We need about five times the number of programs we have now," says Flockhart.

Companies are searching, specifically, for pharmacogenomics directors, who coordinate research programs within the company, bringing together experts from various disciplines, navigating institutional review boards and other obstacles to clinical trials, and coordinating the selection of genes that will be useful in research. "These jobs weren't even known 5 years ago," says Murphy.

Pharmacogenomics is the epitome of interdisciplinary research. It merges molecular biology and pharmacology, two disciplines that have traditionally been far apart. "People with a broad background often are better suited for these positions because you have to consider a lot more" input from various disciplines, says Walter Koch, senior director of the pharmacogenetics department at Roche Molecular Systems, which is a business unit of Roche Diagnostics Corp.

Clinical trial design is another key skill. "One thing that is not widely appreciated by basic researchers is the complexity of clinical trial design. This is a complex science, combining biostatistics, a hypothesis of your mechanism of action, and how you're going to look at clinical endpoints," says Maureen Cronin, who is co-chair of the American Association of Pharmaceutical Scientists' Pharmacogenetics and Pharmacogenomics Focus Group and senior director of technology and development at Genomic Health Inc. in Redwood City, California. "People with a molecular biology background are able to think clinical. That's what we need," agrees Flockhart.

The regulatory angle

FDA has spent nearly 2 years coming up with guidelines for industry on how to handle pharmacogenomics data for approval decisions. Many companies have expressed discomfort with FDA's desire to see pharmacogenomics data, fearing that inexperienced reviewers will draw unsubstantiated conclusions from an early stage test that shows, for example, that a drug causes expression of an oncogene in some percentage of human cells. In fact, many drugs have such effects, so the immaturity of the science casts doubt on the clinical significance of such findings.

For its part, FDA recognizes that uncertainty and is seeking to reassure industry that it will not put undue stress on pharmacogenomics data, but it is also asking companies to voluntarily submit pharmacogenomic data so that the agency's scientists can use it to get up to speed on the science. To that end, FDA has a draft guidance that was open to public comment until February, in which the agency outlines what pharmacogenomic data must be submitted with an application and what data can be submitted on a voluntary basis. The crux of the distinction is that data used to assess the safety or dosing of a drug in clinical trials is deemed vital to the review process and, hence, will be mandatory. But most preclinical data and basic research data will be voluntary.

The guidance has produced a sigh of relief in industry circles, where representatives regard the guidelines as a positive step forward. "I think there will be a little bit of uncertainty, but generally I think [pharmacogenomics data] will fit pretty neatly into categories. [This] is going to require some discussion with FDA, and what they've been saying all along is, 'come talk to us,' " says Brian Spear, director of pharmacogenomics in global pharmaceutical research and development at Abbott Laboratories in Abbott Park, Illinois.

And that's key, because the regulatory environment will have an impact on how enthusiastic pharmaceutical and biotech companies are about adopting the new approach. And that enthusiasm, in turn, is bound to translate into the widespread hiring of scientists trained in pharmacogenomics.

Read the companion article Pharmacogenomics Profiles, also part of this Next Wave feature.

Jim Kling is a freelance science and medical writer based in Bellingham, Washington.