|
|
|
|
Depressing Innovation
in Pharmaceuticals
IBA Update 15 March 04 I wrote this week’s update with Matt Hall, a partner of the Kerdan Group, a strategy consulting firm focused on Life Sciences. You can reach him at mhall@kerdan.com, the firm’s website is www.kerdan.com.===========================
John Morris, Partner, KPMG London
In 2003, many of the world's largest drug companies failed to win U.S. approval for a single new drug. In fact, the U.S. Food and Drug Administration approved only 21 new drugs, marking a steady decline since a peak of 53 in 1996. Yet over that same period, R&D spending by pharmaceutical companies more than doubled, from about $15 billion in 1996 to $32 billion in 2003.
According to numerous managers and researchers, investment in new combinatorial chemistry technology may be one of the major causes for both rising expenditures and the decline in productivity.
These investments are intended to systematize and automate the search for new compounds to create new drugs. Manual screening processes may only generate two or three thousand compound screens per week. But high-throughput screening technologies automatically combine chemicals to test disease efficacy, screening up to 300,000 compounds per day.
Who’s in charge of these high speed processes? Robots, in many cases. First, robotic systems queue up chemical compounds and distribute them into tiny wells on conveyor plates, each containing a different disease-linked target. Positive "hits" from such screenings enable scientists to refine compounds into potential drug leads. Or so the theory goes. So far, however, the results have not borne fruit. A study led by David Newman of the National Cancer Institute concluded that combinatorial-chemistry machines had failed to create a single FDA-approved drug through the end of 2002. As James Hussey, a former Bristol Myers Squibb manager, told the Wall Street Journal in a recent article: Is high-throughput screening the technology underlying future blockbusters? Major pharmaceutical companies appear to think so. Yet industry experts are widely divergent in their viewpoints, from hopeful to pessimistic.
View #1: Understandable growing pains Richard Gregg, vice president of clinical discovery, Bristol-Myers Squibb, in The Wall Street Journal In 1986, Kim Clark and Robert Hayes, professors at Harvard’s Business School, published a seminal study indicating that factory productivity declined for the first several years after installation of significant new technology. As workers and managers became more experienced with the new technologies, efficiency improved at much higher rates than before. Many economists feel that IT investments in the 1980s and 90s followed a similar pattern. Advocates of this "growing pains" view insist that the new combinatorial technologies will follow the same path – an initial decline in output as researchers learn to use new tools, followed by movement to a higher level of benefit. View #2: Expensive Fiasco Other researchers believe that the combinatorial technologies are moving drug discovery in the wrong direction, eliminating the historically important roles played by intuition and serendipity. The development of blockbuster drugs from Penicillin to Prozac has relied on intuition and serendipity. Take the case of statins, the anti-cholesterol drugs now generating $22 billion in annual sales that were conceived by Japanese scientist Akira Endo. Endo had a hunch that fungi could break down cholesterol and then tested 6000 different kinds of fungi by hand before finding one that did so. Another noted scientist, Arvid Carlsson, Nobel laureate for research on neurotransmitters, recently told The Wall Street Journal: Combining robots and intuition It’s hard to manage and predict the results of processes that require serendipity. This is an institutional problem for large drug companies and their investors, who would like reliable estimates of future growth and earnings. Companies like Bristol-Myers and Pfizer are working to integrate the intuitive and the systematic, supplementing the intuition of their research scientists with the speed and efficiency of their high throughput screening technologies. The Economist concluded a recent survey of new drug discovery technologies by noting: This seems like a logical step, and it will undoubtedly aid in the discovery of many new molecules. But it may not help in the creation of new blockbusters, because many of the historical paths to blockbuster products have been meandering ones, while high throughput screening conceptualizes the drug discovery process as a linear sequence. Based on what we know about breakthrough innovation, it is quite possible that combinatorial technologies are much better suited for extending existing drug products than for developing entirely new classes of drugs. If this is the case, then large drug companies are actually reducing the effectiveness of their most productive source of new blockbusters – the random connections made by well-trained researchers. Lessons from other industries While intuition and serendipity may be difficult to manage, there are well-tested techniques for tackling problems, such as blockbuster drug development, that require invention. One method, pioneered by NASA in the 1960s, is to run two independent project teams with the same goal and different approaches. These teams meet occasionally to compare results, but they approach a problem using fundamentally different research paths. With the costs of development now approaching a billion dollars per drug, isn’t it surprising that more drug companies aren’t pursuing parallel approaches in their quest for new blockbuster products? Robotic drug development may be more efficient at screening compounds, but has yet to prove it will yield the breakthrough innovations that companies need for growth and profitability. More Information |
|
All content on this site licensed under a Creative Commons License.
|
