Pharma Facing Tough Challenges, Speakers Say at AAAS S&T Forum

William Haseltine

The pharmaceutical industry, while still very profitable, is facing significant challenges, with its ability to translate new ideas into products threatened and questions continuing to swirl around the process by which drugs are approved, according to speakers at the recent AAAS Science and Technology Policy Forum.

They spoke of companies that have grown too large and too focused on developing "blockbuster" drugs that can be marketed to the widest possible patient population. They spoke of the conflicts of interest that can arise when industry funds 80 percent of the clinical trials conducted in the United States. But they also spoke of the long-term promise of biomedical innovation, and urged careful analysis of the underlying statistics on drug company productivity.

"What I see is an industry in very deep trouble," said William Haseltine, chairman of Haseltine Global Health, LLC, and the founder of Human Genome Sciences Inc. The number of new drug approvals per year is down substantially, he said, despite a 10-fold increase in spending by traditional pharmaceutical research companies.

Biotechnology companies, with only about one-fourth the R&D; spending of the large pharmaceutical companies, have been effective in getting new products into clinical trials, Haseltine said. But they have had limited success in winning prompt regulatory approval.

In 2006, there were 418 biotech products in final, phase III clinical testing, according to John Engel, managing partner of the Engel & Novitt law firm. Only four such biologic products won approval that calendar year, however.

Scott Stern

Still, the productivity crisis in the biopharmaceutical industry "is not as dire as sometimes claimed," cautioned Scott Stern, an associate professor at Northwestern University's Kellogg Graduate School of Management. He said the industry is going through a long-term change from chemistry-based products to those based on biology. Such technological change—often disruptive and costly—can take years to diffuse into the marketplace, he said. While the rate of drug approvals slowed from 2000 to 2005, Stern said, there has been a sharp increase in the number of promising compounds at earlier stages of the drug-approval process.

The cost of bringing a new prescription medicine into the marketplace is substantial, estimated at $802 million by a 2001 Tufts University study. Another Tufts study last year pegged the cost for a new biopharmaceutical at $1.2 billion. But Stern said the cost squeeze may not be as grim as sometimes portrayed.

Development costs are rising, he said, but sales also have experienced a dramatic growth in recent years, due in part to some blockbuster drugs with wide use and increasing revenues from applications discovered after drugs are initially approved for the market.

While Stern was optimistic about the long-term prospects for drug discovery, there remain serious ethical and procedural questions about the clinical trials that drugs must undergo before they can be approved for the marketplace. At each step in an industry-sponsored drug trial—from design of the study to recruitment of volunteers to analysis and interpretation of the data—there can be problems, according to Dr. Deborah Zarin, director of ClinicalTrials.gov, a registry of clinical trials at the National Library of Medicine.

Dr. Deborah Zarin

Companies can find ways to structure drug trials so that they maximize the prospects for a favorable outcome, Zarin said. She noted a case in which a proposed pain medication was linked to an increase in heart attacks during clinical trials when compared to an existing pain medication. The company switched to comparing it instead with a more risky drug in its class. Using such tactics, she said, "You can usually figure out a trial that gets to 'yes.' "

The conduct of trials also can be compromised by patient recruitment tactics, Zarin said. Sanofi-Aventis recruited patients for its trial of the antimicrobial agent Ketek by paying some 1,800 physicians as much as $400 per patient enrolled, she said. Food and Drug Administration inspectors found evidence of fraud, including fabrication of patient enrollment data.

There also can be issues at the publication stage, Zarin said. A journal article by industry-funded investigators found that the drug Paxil was effective in treating major depression in adolescents. But another analysis concluded it was ineffective and perhaps risky. It also turned out there were two unpublished studies that showed no efficacy, she said.

In another case, the Journal of the American Medical Association accepted an article claiming positive results for Celebrex, an anti-inflammatory drug, after six months of clinical testing. Only after publication, Zarin said, did the editors learn that the length of the trial was 12 months, and the 12-month results (which showed no statistical difference between Celebrex and another drug) were available at the time of publication. (In 2005, the FDA told Pfizer, the drug's maker, to put a "black box" warning on the Celebrex label to note the drug may increase the chance of heart attack or stroke.)

"So journal editors can't validate all the data and analyses," Zarin said. "They don't know if the data are complete and correct, and they don't know if the data are analyzed with the appropriate methods."

Trial registries, such as the ClinicalTrials.gov site that Zarin manages, can be helpful. Her registry provides summary information on the testing protocols being used in a trial and links to publications that result from the study. But it does not provide all of the source data from the completed trial. The registry is adding about 220 trials per week from around the world, Zarin said, with more than 40,000 trials now registered.

Leading medical editors have made participation in the registry a requirement for those seeking to publish results in their journals. Still, Zarin said, even with registries, there are many trials for which full results are not presented. While there are proposals in Congress to require that study results be entered into registries, Zarin said, none of the bills would mandate independent access to the source data or the entire study protocol.

"So, for example, it would have Merck entering the results on Vioxx," Zarin said. "It would have Pfizer entering the results on Celebrex. We wouldn't know if the data were accurate or if there were an error." (Vioxx, an anti-inflammatory drug of the same type as Celebrex, was withdrawn from the U.S. market in 2004.)

Shannon Brownlee

Shannon Brownlee, a fellow at the nonprofit New America Foundation, said the billions of dollars spent on clinical trials can lead to conflicts of interest and subtle effects on the outcomes of the research. She cited a survey in the New England Journal of Medicine of 70 studies on the safety of blood pressure drugs called calcium-channel antagonists. Of those studies whose authors were supportive of the drug, 87 percent had received funding from the company compared to 20 percent of those who were critical of the drug. Another large survey, involving 1,140 original research studies, found strong and consistent evidence that "industry-sponsored research tends to draw pro-industry conclusions."

Brownlee also cited a study last year in the American Journal of Psychiatry of 30 head-to-head trials involving second-generation anti-psychotic medicines. In 90% of the studies, the outcome favored the sponsor's drug. If a drug did well against a competitor in one study, it came in second when the study was funded instead by the competitor.

Clinical trials also can be biased when researchers fail to publish negative results, Brownlee said, or try to gloss over negative findings. In one study, she said, researchers reported that "most adverse events were not serious" when, in fact, seven children had to be hospitalized for side-effects.

Brownlee called for more federal funding of clinical trials and establishment of an Institute for Clinical Evaluation to help assess the quality of clinical trials.

Haseltine, while acknowledging that there have been problems with the conduct of some drug research, said the industry faces deeper underlying problems. "It's not just a question of output from research funding, or problems in the drug approval process," Haseltine said. If clinical trials can be maneuvered and twisted, he added, the drug companies "are not twisting them very effectively." Even if people are tilting the playing field, he said, the overall approval rate is not very good.

What's going on? Haseltine argues that the large pharmaceutical companies are "simply too big." They often are looking to develop products aimed at the broadest possible market and with the largest impact on the bottom line. "You develop drugs for a market, not a disease," Haseltine said. "And that is a fundamental mistake." He cited the case of Celebrex. Haseltine said it is an appropriate drug for a small class of patients who would benefit from it. But he said some drug companies are pursuing what he calls a "reverse Cinderella" strategy by trying to "put a small foot into a big shoe."

The net result, he said, is that the industry's ability to translate ideas into new products is "deeply imperiled and threatened." One solution would be to separate the marketing function from the R&D; function, Haseltine said. Company officials need to keep very close to the development process. "We need experts with an intuitive, obsessive feel for the science and the disease," he said.

Earl Lane

17 May 2007