The business of pharmacogenomics

Although the exact cost is in dispute, Washington, DC-based Pharmaceutical Research and Manufacturers of America estimates that this cost currently averages $500 million. This increasing R&D cost is due to more than inflation; drug development has become more complicated. According to pharmaceutical industry analyst Hemant Shah of HKS and Co., many of today’s drug targets are “far more difficult, far more complex, and far more uncertain” than ever before.

Compared to 20 years ago, a growing proportion of drugs is prescribed for chronic conditions rather than as short-term antibiotics or painkillers. Roger Perlmutter, executive vice president for basic research at Merck Research Laboratories, notes that therapies aimed at long-lasting conditions such as hypertension, high cholesterol, diabetes, or neuropsychiatric diseases are usually going to be taken daily for years. Therefore, they must be free of side effects that in short-term prescription medications might be tolerated. In addition, older patients who take many of these drugs are also taking other medications, so the issue of harmful drug interactions is an important concern.

Taking these factors into account has increased drug development costs, while also making it more difficult for candidate drugs to survive clinical testing. For example, the U.S. pharmaceutical industry continues to increase the number of compounds that it studies. Between 1995 and 1999, the number of compounds under development increased by 35% to 7434; however, the number of new product launches declined 22% to 56 over the same period.

Drug R&D program size
Opinions regarding the advantages and disadvantages of very large drug development programs have long been divided. Samuel Isaly, a pharmaceutical industry analyst and partner in the New York investment firm OrbiMed Advisors, comments, “Historically, it [drug R&D program size] hasn’t mattered.” He believes that success in drug discovery has depended on innovative ideas rather than large research facilities and a lot of money.

This opinion is consistent with the findings of a Tufts University study, Pharmaceutical Industry Innovation Is More Dispersed Despite M&A Activity, by Joseph A. DiMasi (Tufts Center for the Study of Drug Development). Results indicated that the share of FDA new drug approvals won by the four largest pharmaceutical firms declined from 30% in the 1960s to 18% in the 1990s. A similar trend exists when considering the eight largest drug companies. “Logic and experience might suggest that increased M&A activity should lead to a greater concentration of drug development activity among fewer firms, but the opposite has happened,” said Kenneth I. Kaitin, director of the Tufts Center. “We’ve found that pharmaceutical innovation has become increasingly diffuse across firms, particularly as the ranks of the research-based pharmaceutical industry expanded during the 1990s.”

The Tufts study indicates that the number of firms obtaining approvals for new drugs increased 84% from the 1970s to the 1990s. Of the 50 firms with a single new drug approval during the 1990s, for 41 firms, this was their first-ever new drug approval. Many of these are startup firms, often biotechnology companies. Kaitin said that advances in biomedical science may have fostered less concentration of new drug development among existing firms and stimulated new entry to the industry.

R&D restructuring
Kim Slocum, director of strategic planning and business development at AstraZeneca, is concerned that mergers may actually decrease R&D productivity by creating larger, more bureaucratic R&D organizations. Steve Arlington, head of pharmaceutical R&D consulting at PricewaterhouseCoopers, suggests, “Those companies which fail to restructure their R&D processes, track what they are doing, implement the right IT strategies, and put the right skills in place will not survive in their current form.” Three companies substantially enlarged by recent mergers and takeovers—Aventis, GlaxoSmithKline (GSK), and Pfizer—recently have restructured their R&D programs.

Aventis, created in 1999 by the merger of the life sciences operations of Hoechst and Rhône-Poulenc, has an annual R&D budget of about $2.8 billion. The firm renamed its R&D unit “Drug Innovation and Approval” to emphasize the need to develop innovative new drugs and obtain regulatory approval as soon as possible. To promote this goal, researchers working in different departments but having related projects have been placed under a global management. The objective is to optimize compound leads more quickly through greater collaboration and critical evaluation. To help maintain a commercial focus, sales and marketing staff members, as well as researchers, review projects to assess their commercial potential.

GSK is extensively restructuring its R&D program after the 2000 merger of Glaxo-Wellcomeand SmithKline Beecham. GSK’s drug R&D program is massive—approximately 15,000 scientists and an annual budget of about $3.3 billion.GSK plans to split its R&D effort into six separate units. Each will focus on a different therapy area and compete for corporate R&D funding. According to GSK research head Yachi Yamade, the objective is to develop small entrepreneurial business cultures within its large R&D organization, thereby boosting research productivity, while still enjoying the advantages of scale provided by a large R&D organization. David Beadle, a pharmaceutical industry analyst at UBS Warburg, said the new GSK structure is encouraging because it means that research funding will be more commercially oriented.

The proposed new GSK R&D structure appears similar to the decentralized model recently announced by Pfizer chief executive Henry McKinnell. Roche already has organized its R&D into semiautonomous units, while trying to maintain cross-fertilization of ideas through frequent travel and meetings among researchers. It has also established an autonomous subsidiary to run some of its clinical trials.

R&D collaborations have become increasingly important in new drug development. Companies with large R&D budgets can fund a larger number of collaborations with outside organizations, bringing their creativity to bear on large internal R&D programs. For example, George M. Milne, Jr., senior vice president and president of worldwide strategic and operations management, said that Pfizer, with a 2001 R&D budget of approximately $5 billion, has more than 450 partnerships with external collaborators. These collaborations are enabling 12,000 Pfizer researchers at six sites to work more effectively on 156 drug development projects in 19 disease areas.

Big is not always enough
Even very large R&D programs do not guarantee a steady flow of profitable new medicines from a firm’s drug development pipeline. For example, recently, GSK was forced to withdraw its irritable bowel syndrome drug Lotronex from the market and terminate development of its experimental drug Enrasentan after it showed no benefit in intermediate Phase II clinical trials for chronic heart failure. The FDA has rejected the antibiotic Factive and cancer and heart drugs. Release of the asthma drug Advair has been delayed. To quickly revive its depleted new product pipeline, GSK plans to acquire smaller firms with promising drugs in late-stage development and enter into drug comarketing agreements with rivals commercializing promising drugs.

Currently, the way to beat the competition and compensate for declining sales of drugs going off patent and facing generic competition is developing new blockbusters, drugs with annual sales in excess of $1 billion. But could this be changing?

According to Isaly, “We are more and more of the opinion that size begins to matter in drug discovery.” He believes the combination of genomics and chemistry to create major drug candidates is changing the industry. Genomics holds the promise of identifying many new drug targets. Drug development chemistry increasingly requires extensive automation of very repetitious research to identify a few prime candidates among many possibilities.

In contrast to Isaly’s view, some experts suggest that the genomic reinvention of drug development may change the pharmaceutical industry landscape from a handful of big companies with blockbuster drugs to a host of small and midsized companies with targeted, smaller market medicines. The pharmacogenomic approach to drug development first identifies target genes and then designs drugs to turn them on or off to halt or reverse a disease process. Wolfgang Sadée (University of California, San Francisco) suggests that the role of genetic variability in disease and therapy means that smaller, genetically defined patient populations can be treated more effectively with tailored drugs than with a single blockbuster drug. The net result, besides being a market for this targeted drug, is a smaller market for the blockbuster.

Global scale needed
To develop targeted drugs, genomics companies need to study large groups of people on a global scale, those with and without a particular disease. For example, Perlegen Sciences, a new subsidiary of Affymetrix, creators of DNA-scanning technologies, can now read 50 individual genomes, identify the millions of genetic variations among individuals, and find patterns in those variations. “Those 50 genomes will yield 3 to 4 million common genes. What we’re looking for is the 10 to 20 genes that together cause a disease,” explained Brad Margus, Perlegen CEO. “To find those 10 to 20 genes, we’ll need to look at 1 or 2 million people with disease at various places in the world.”

DNA Sciences has established a data bank called Gene Trust for individuals and families with inherited disorders to contribute blood samples for study. The hope is that, by having a relatively large population of people with comparatively rare diseases, the genetic cause of the disease can be determined and a drug developed. Of course, the market for the genomic drug may be relatively small. This could be a real issue in drug pricing if a means cannot be found to reduce the costs of drug development.

In implementing genomics, John Niblack, president of Pfizer Global Research and Development, still sees an advantage to having a very large R&D program. He said, “As anemerging understanding of the genome expands our ability to hypothesize a variety of potential medicinal applications for new molecules and mechanisms, the span of research expertise and involvement of the new Pfizer permits rapid, aggressive exploration of those possibilities.” Milne noted that as a result of investments in leading technologies, Pfizer has achieved severalmajor productivity advances. These include a projected 40-fold increase in its ability to conduct high-throughput screening of compounds, from generating 3.8 million data points in 1998 to a projected 150 million data points in 2002. These advances are already generating new and better leads in a range of therapeutic areas, according to Milne, and could enable Pfizer to more rapidly exploit leads identified by pharmacogenomics.

Drugs that fail during clinical trials because of unpredictable patient responses cost the pharmaceutical industry billions of dollars annually. Often drugs are denied FDA approval because a relatively small patient population reacts adversely to them. Allan Roses, worldwide director of genetics at Glaxo-Wellcome said, “Selection of predicted responders offers a more efficient and economic solution to a growing problem that is leading governments and health care providers to deny effective medicines to the few because a proportion of patients do not respond to treatment. The economy of predictable efficacy, limited adverse events, lower complications owing to targeted delivery, and increased cost-effectiveness of medicines will improve health care delivery.”

Ronald M. Norton, director of technical services at PPGx, notes, “DNA tests based on genetic variations can predict how a patient will respond to a particular medicine. Clinicians will use them to select optimal therapy and tailor dosing regimens; the benefits will include reduced incidence of adverse drug events, improved clinical outcomes, and reduced costs.”

Currently, many small biotechnology and pharmaceutical companies are reluctant to take the risks and assume the high costs of clinical trials. As a result, they often sell joint-development rights to candidate drugs to large firms. Reducing these risks through pharmacogenomics would increase the willingness of small firms to go it alone in drug development. However, comarketing agreements to enable a small company’s innovative drug to take advantage of the large sales forces maintained by major drug companies could still be common.

Large or small?
In conclusion, opinion appears divided on whether large or small drug development programs are inherently more efficient in developing new drugs. The findings of the Tufts University Center for the Study of Drug Development do appear to indicate that small R&D programs can often do well in competing with very large R&D organizations. If combinatorial methods have indeed increased the advantage of large R&D programs, genomics could well have the opposite effect. Only close study of various firms’ R&D programs will determine whether large or small programs are more effective in today’s changing research environment.

John K. Borchardt is a science writer based in Houston. Send your comments or questions regarding this article to mdd@acs.org or the Editorial Office by fax at 202-776-8166 or by post at 1155 16th Street, NW; Washington, DC 20036.