May 2008’s edition of Harvard Business Review has an article titled ‘Rebuilding the R&D Engine in BigPharma’. In this article Jean-Pierre Garnier -CEO of GlaxoSmithKline- talks about the changes that BigPharma’s R&D process has to face to stay competitive in the new market.

See if you find anything familiar in these quotes:

The most significant [of the actions] is the break-up of our formerly mammoth R&D organization into small cross-disciplinary groups, each focused on a family of related diseases.

Indeed, the share of experimental drugs that fail in the clinical stage (when testing on people occurs) has actually risen [since the genome project], hitting a record 93% in 2006 […]

[…] pharmaceutical companies have maintained an organizational setup that worked will in the 19060s: a pyramid with functional silos […] all joining at the top. […] the pyramid contained only a few management layers; there were fewer projects; scientists worked together on one campus; and resource allocation was relatively straightforward. [on the new century] The pyramid became a monster and everything suffered. Silos could not communicate seamlessly. Overly complex matrix teams were created to try to overcome the rigidities. the decision-making process slowed. And middle and upper management lost their command of the fast-evolving science.

[R&D organizations] have traditionally promoted their best scientists to management positions - sometimes without paying enough attention to leadership abilities. Complexity and the leadership void have given rise to teams that focus too much on process and too little on producing meaningful results[…]

Traditionally, new drugs with potential for treating a common disease […] have been testes on a large, diverse patient population in clinical trials. If the trials produced the expected results and the FDA approved the drug, the result was an "instant blockbuster. However, it is common for a drug to cause side effects in a small segment of the target population. Years ago these side effects would often go undetected until after the drug was on the market and millions of patients had used it. Today the FDA requires companies to try to identify [those] before a product launch; this has led to a record number of initial rejections and [delays]

A solution would be to begin by limiting the clinical trials to a highly uniform segment of patients […] I testing produced good results, the company could seek the FDA’s approval for marketing the drug to that segment only. […] Development of the drug should continue in the same way: A second uniform patient segment should be chosen. And so on.

[this strategy] would address society’s low tolerance for surprises […] and should simplify and speed up product development. Ultimately, everyone wins.