Oct. 16, 2006 | Innovation is a major component of the pharmaceutical industry - the key to discovering new drugs, improving drug safety, and reducing costs to market. Innovation is also accompanied by 'myths,' such as in individuals are responsible for the success of such innovations. It is widely held that Tim Berners-Lee created the World Wide Web, yet his invention was highly dependent on pre-existing Internet standards. The ubiquity of the Web took off because developers not only used his code but also extended it and set up thousands of Web servers. Finally, he did not create the content on the Web, which for most of us "is" the Web!
In drug discovery, taxol was first identified in the mid-60's from pacific yew trees, but it took the extracts from several 100-year-old trees to produce enough drug for a single patient. It took many researchers many more years to efficiently synthesize the compound and elucidate its mechanism of action - essential steps in realizing the value of taxol, which is widely used in the treatment of cancer.
In both examples, access to relevant information was critical to enabling the next series of gains. If anything obstructed information access, the process quickly becomes protracted, and the chance of success plummets. In many cases, more benefits can be achieved by improving the flow of information rather than holding to it locally. This is currently taking place in many media venues proliferating in the "social Web", such as blogs and public tagging spaces.
Drug companies build a lot of value around their IP, but this works best when the IP is directly linked to the product being marketed. If information is withheld around a necessary process, the value of which is not directly tied to the consumer, then this often produces inefficiencies in R&D, paradoxically restricting the value of the information. In drug R&D, IP is most often associated with chemical substances and processes. On the other hand, safety is associated with a set of measurable and reproducible criteria. Assessing safety is a necessary procedure based on a common standard set of facts and guidelines, but it is not defined by the industry. Hence it is difficult to reconcile this common information requirement with company-centric IP. There is a need to objectively analyze and determine information value, and to come to some form of agreement on placement of information in a Commons.
Such information does not arise for free - it is produced through the investment of a company's time and money - so why give it away for free? Surely there is more value gained by combining the knowledge than lost by sharing it (even between competitors), since a complete understanding is more useful than a partial one. Consequently, the inability to re-use all information forces researchers to spend more time and money on re-acquiring knowledge. What's more, if all current information could be combined and used, new drugs would typically be safer and with fewer late stage failures - this is where a Commons could have an enormous impact on the industry. If all companies are facing the same problem in safety, and no one is gaining much from their own safety and toxicity signals, then much can be gained by sharing.
Degree of Access
Another key issue concerns the degree of access to information in a commons. Most R&D activities rely heavily on published, copyrighted research, yet whose content can be used develop products. Access rights to such information, including supporting data, has a major throttling effect on subsequent innovation. This is also true for information generated internally as part of drug R&D: if more of the contributions of others were available, it could be effectively used to promote innovation upstream and downstream. Clearer definition of access rights and how to find them can greatly accelerate rights negotiations, saving researchers from time spent on entangled legal discussions. ScienceCommons has been working to establish these definitions and practices, in a way that brings the legal system into the Age of the Social Web.
The notion of Research Commons where both experimental evidence and insights are "posted" to a community is not new. Organizations such as DiscoverLife (www.discoverlife.org), Global Information Commons for Science Initiative (www.codata.org/), BIRN (www.nbirn.net), NeuroCommons (www.neurocommons.org), Alliance for CellSignalling (www.afcs.org) and others support a research commons approach for its members. A common feature is the importance in attribution or authorship in posted works and a sufficient amount of provenance and other forms of scientific metadata. This includes the continued addition of new links and quality annotations on commons entities, whereby the information remains fresh and aggregates new knowledge onto legacy data. This is the nature of social Web systems such as blogs and media tagging sites. Semantic Web standards would further improve the quality of the annotations used, and ensure compliance with controlled vocabularies.
Once it is possible to discern which information has greater value in a commons space then it would be quite straightforward to develop an agreement of the policies and programs for all drug R&D and biotechnology companies. Since agencies like the FDA are requesting more safety and toxicity information be produced and submitted in a standard way by drug manufacturers, they may be able to help define what set of information has more value in a drug safety
commons. In some regard, they are the primary consumers of the safety information, yet they also hope to use the information not just for new drug approvals, but to improve the overall industrial R&D process through the re-use of knowledge. For example, a 50 percent reduction in failures in Phase 3 clinical trials would have a potential savings of ~$300 million per new drug.
Once a Drug Safety Commons has been established, a variety of derivative applications will be developed that can access as well as update the information in organized ways and used by industry researchers internally. The innovative re-use of a more complete set of information - including early toxicity assay development, mapping in vitro and animal studies for better prediction, re-analysis of clinical data and samples - then becomes the new, more effective driver for drug R&D innovation. Such community-derived innovation will resemble the spirit of today's Web applications such as Google, del.icio.us, flickr, and others, yet the scale of the biomedical benefits realized by these new forms of R&D practices will be uncommonly profound.
Eric K. Neumann is senior director product strategy at Teranode. E-mail: eneumann@teranode.com.
Subscribe to Bio-IT World magazine.