The Atlantic Home

This week, Merck announced a $41 billion offer for Schering-Plough. Last month Pfizer agreed to purchase Wyeth for $68 billion. Roche is negotiating to buy the part of Genentech it does not already own. Rumors of other impending mega-deals swirl. I believe that this current wave of mergers reveals a disease at the heart of the pharmaceutical industry.

The fact that so many companies are now merging reflects the failure of each company to discover and develop its own replacement pipeline. To maintain growth, a pharmaceutical company must either produce enough new products to replace those that have gone off-patent or acquire rights to distribute drugs created by others. This is clearly not happening on a large enough scale. A small number of patented drugs, each with annual sales of $1 to $5 billion, accounts for most of the profits of the large pharmaceutical companies, and these profits are vanishing as the patents expire. Meanwhile, the number of new drugs approved for sale annually has steadily decreased over the past 15 years.

What has gone wrong? Are today's unsolved medical problems more difficult than those of previous generations? Are regulatory barriers higher? Such is the common wisdom.  However, my 30 years of experience as a consultant, and as the founder and CEO of several biotechnology companies, suggests a different primary problem: the marketing-driven demands of the global giants.

Consider the following paradox: despite the declining productivity of the pharmaceutical industry, the past 20 years have been a golden age for biomedical research. We have decoded the entire human genome as well as those of most of the viruses, bacteria, and parasites that have plagued us for centuries. Our ability to understand the differences between normal and disease states has increased exponentially. New opportunities to treat hitherto intractable diseases such as Alzheimer's, cancer, and heart disease abound. Advances in imaging permit us to detect diseases that were previously hidden, often at a stage early enough for life-saving intervention. Computer models have given us powerful new tools for linking biological causes, disease effects, and probable consequences of specific treatments, while computer-driven robots allow us to sort though libraries of new drug candidates at a vastly accelerated rate. In fact, the research departments of each large pharmaceutical company produce hundreds of ideas for new drugs each year. Why, then, the slowdown in bringing these drugs to market?

I believe that it is not a failure of science but rather the context in which such science is practiced. The process of drug development is long and arduous, and throughout the entire process, a candidate is continually evaluated for market potential. At large companies, products that are technically promising are terminated if the marketing potential is thought to be too small. And the height of that market hurdle has risen as the profits of the large companies have grown. Today, programs that are thought to have an annual sales potential of less than $1 billion are usually stopped in their tracks. Some companies have abandoned their work in entire areas of medicine, such as antibiotics, because they believe the markets are too small to make a difference to their total sales.

This emphasis on the bottom line even influences the way clinical trials are designed. Companies are in a position to choose how each new drug is deployed, and drugs that are approved for limited use have limited market potential. Therefore, most clinical trials are designed to optimize sales, not to optimize the chance that a new drug will be approved for its most effective, if limited, use.

The results of this strategy are evident. Fewer than one in 100 new ideas reach clinical trials, and fewer than 10 percent of these are approved for sale. In the meantime, 10 to 12 years have typically elapsed between discovery and approval, consuming half the twenty-year patent life of each drug candidate. And when drugs are approved, problems often arise in the quest for ever-larger markets. In worst-case scenarios, the side-effects of approved drugs--perhaps tolerable in small treatment populations--are inappropriately minimized as the product is brought to mass markets. In one such instance, the drug Vioxx, approved as painkiller rather than a treatment for a smaller group of aspirin-intolerant patient, had to be withdrawn five years after it was approved by the FDA.

Mergers will not solve any of these underlying problems. In fact, as the size of the companies increase, the long-term problems will become worse.

Many large companies have looked to biotechnology for solutions. For years, the hope has been these new companies would produce enough new drugs to meet both their own needs and those of the large of the pharmaceutical companies. But despite such obvious success stories such as Genentech, Amgen, Biogen, and Gilead, recent history has proven otherwise. Most biotech companies lack the management expertise and capital required to bring new drugs to market. At some stage, they hand over development and marketing to established organizations that then apply the same market-driven selection criteria to the biotech products as they do their own. New drug candidates that don't meet the market requirements are terminated.

The solution is clear: research and development of new drugs should be separated from marketing organizations. The cost of drug development must also be dramatically reduced. The electronics and computer industries have shown us how this can be done: new products are initially invented and developed by small companies that provide innovation and judgment. These small companies in turn establish well-defined, process-based contracts with service companies. Once developed, the products are then sold by global marketing organizations.

It is time to do the same for the early stages of drug development. During the past decade, more and more process-based companies--mostly located in Eastern Europe, India, and the Far East--have arisen to provide specialized services for each step of the drug-development and clinical-trials process. At the same time, a new generation of biotechnology companies has arisen to provide the innovative science and judgment necessary to guide product development. Their semi-virtual design processes may allow many more drugs, each developed for its optimal use, to reach the market. Once approved, some of these new drugs may even become blockbusters--the initial sales projections for most of today's best-selling drugs were small, and by today's standards, most would have never made it to the market.  

In summary, the current wave of mega-mergers does not solve but rather exacerbates the deeper problems of the pharmaceutical industry. Despite the industry's appetite, its emptying product supply will never be filled without a reliable research-and-development engine. Radical restructuring, not merger mania, is the need of our time.