Knowledge Networks and Markets in the Life Sciences by OECD

VIEWS: 22 PAGES: 131

									Knowledge Networks
and Markets in the
Life Sciences
 Knowledge Networks
and Markets in the Life
This work is published on the responsibility of the Secretary-General of the OECD.
The opinions expressed and arguments employed herein do not necessarily reflect
the official views of the Organisation or of the governments of its member countries.

This document and any map included herein are without prejudice to the status of
or sovereignty over any territory, to the delimitation of international frontiers and
boundaries and to the name of any territory, city or area.

  Please cite this publication as:
  OECD (2012), Knowledge Networks and Markets in the Life Sciences, OECD Publishing.

ISBN 978-92-64-11937-6 (print)
ISBN 978-92-64-16859-6 (PDF)

The statistical data for Israel are supplied by and under the responsibility of the relevant
Israeli authorities. The use of such data by the OECD is without prejudice to the status of the
Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of
international law.

Photo credits: Cover © vladgrin -

Corrigenda to OECD publications may be found on line at:
© OECD 2012

You can copy, download or print OECD content for your own use, and you can include excerpts from OECD
publications, databases and multimedia products in your own documents, presentations, blogs, websites and
teaching materials, provided that suitable acknowledgement of OECD as source and copyright owner is given.
All requests for public or commercial use and translation rights should be submitted to
Requests for permission to photocopy portions of this material for public or commercial use shall be addressed
directly to the Copyright Clearance Center (CCC) at or the Centre français d’exploitation du
droit de copie (CFC) at
                                                                                   FOREWORD –   3


            This report introduces the concept of knowledge networks and markets
       (KNMs), and discusses the new organisations and mechanisms that are
       emerging to share and to trade an increasing variety of knowledge-intensive
       assets. It describes numerous case studies of such initiatives in order to illustrate
       the variety of open knowledge management approaches. The report discusses
       the technological, economic and industrial environments that have led to the rise
       of KNMs, and in particular delves into advances in both computer science and
       knowledge valuation that could further facilitate the representation and
       exchange of knowledge assets. The report argues that the creation of such
       exchange mechanisms is an important new trend in the life sciences, and
       particularly in the health sector, with potentially profound influence on the
       innovation process. Despite the novelty of KNMs, this report identifies some
       early policy lessons about the role of governments in the creation and
       maintenance of KNMs.
            As the bedrock for subsequent analysis, an OECD workshop on knowledge
       markets in the life sciences was held in Washington, DC in October 2008 to
       discuss the nature of these new exchange mechanisms, the forces that are
       driving their creation, and their implications for the innovation process. It was
       one of the first workshops dedicated to understanding the potential of open
       science and open innovation approaches in the life sciences. Experts from a
       cross-section of relevant fields – including information technology (IT) and data
       management companies, pharmaceutical and biotechnology firms, public
       research organisations, government funding and regulatory agencies, technology
       transfer groups, clinicians, and patient organisations – participated in the work.
       The aim of the workshop was to help policy makers understand the importance
       of new knowledge networks and markets and the role that policy can play in
       facilitating their emergence. Many of the case studies in the report were based
       on discussions that began at the workshop.
           Since that initial workshop, the OECD’s Working Party on Biotechnology
       has carried out significant substantive analysis of the most recent research and
       debates around the growth and impacts of KNMs. The synthesis of that work is
       presented here.
           The report was drafted by Benedicte Callan and Iain Gillespie, under the
       direction of the Working Party on Biotechnology. Special thanks go to Kate
       Hoyle at the University of Toronto, whose rapporteur's summary of the
       Washington workshop provided an invaluable starting point for our analysis.

                                                                                                       TABLE OF CONTENTS –           5

                                                 Table of contents

Executive summary............................................................................................................ 9
Chapter 1. The rise of knowledge networks and markets as enablers of
open innovation .............................................................................................................. 13
   Definition of knowledge networks and markets (KNMs) ............................................. 15
   Varieties of KNMs ....................................................................................................... 17
   A diversity of structures ............................................................................................... 23
   Open questions about KNMs ....................................................................................... 24
   Report structure ............................................................................................................ 26
   Notes ............................................................................................................................ 28
   References .................................................................................................................... 29
Chapter 2. Knowledge flows ......................................................................................... 31
   The definition of knowledge ........................................................................................ 34
   The new knowledge complex....................................................................................... 36
   Biopharmaceutical industry perspective on knowledge flows ..................................... 37
   IP market failures and knowledge flows ...................................................................... 40
   Conclusions .................................................................................................................. 42
   Notes ............................................................................................................................ 44
   References .................................................................................................................... 45
Chapter 3. Advantages of knowledge networks and markets .................................... 47
   From open innovation to knowledge networks and markets in drug development ...... 48
   A network of pharmaceutical consortia ....................................................................... 51
   The promise of information technology and knowledge markets in the life
   sciences ........................................................................................................................ 55
   Conclusions .................................................................................................................. 56
   Notes ............................................................................................................................ 59
   References .................................................................................................................... 60


Chapter 4. Theories for building knowledge networks and markets ........................ 61
   Typologies for knowledge networks and markets........................................................ 65
   Data mining in biomedicine ......................................................................................... 67
   Conclusions .................................................................................................................. 69
   Notes ............................................................................................................................ 70
   References .................................................................................................................... 71
Chapter 5. Case Studies of knowledge networks and markets .................................. 73
   Varieties of knowledge networks and markets ............................................................ 75
   Online auctions, exchanges and brokers ...................................................................... 93
   Conclusions .................................................................................................................. 94
   Notes ............................................................................................................................ 96
   References .................................................................................................................... 98
Chapter 6. The importance of knowledge valuation for knowledge networks and
markets ........................................................................................................................... 99
   Measuring and reporting intellectual assets at the firm level ..................................... 101
   Exploiting intellectual assets to create national wealth .............................................. 105
   Leveraging knowledge to create equity markets ........................................................ 108
   Conclusions ................................................................................................................ 110
   Notes .......................................................................................................................... 111
   References .................................................................................................................. 111
Chapter 7. Conclusions and research needs in knowledge networks and
markets ......................................................................................................................... 115
   What are knowledge networks and markets (KNMs)? .............................................. 116
   What benefits might knowledge networks and markets offer? .................................. 117
   What policy support do KNMs require? .................................................................... 119
   What should be the focus of future work? ................................................................. 121
   Future research programme........................................................................................ 121
   Note ............................................................................................................................ 123
   References .................................................................................................................. 124
Bibliography ................................................................................................................. 125

                                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                                                TABLE OF CONTENTS –         7


Box 2.1.        Examples of collective systems (i.e. KNMs) for innovation in the
                biosciences .............................................................................................. 37
Box 6.1.        Case studies in biomedicine .................................................................. 102
Box 6.2.        Commission on Intellectual Capital Framework ................................... 105
Box 6.3.        Exploiting intellectual assets to create national wealth ......................... 107
Box 6.4.        Structured finance facility – SecurePharma .......................................... 109


Figure 2.1.     Contributions to growth in global R&D, 1996-2001 and 2001-06 .......... 33
Figure 2.2.     Growth in numbers of patents co-owned between businesses................. 34
Figure 2.3.     Percentage of new approved drugs based relying more than 50%
                on externally derived technology, 1989-2004 ......................................... 38
Figure 2.4.     Growth in receipts from international licensing of patents ..................... 39
Figure 3.1.     TI Pharma structure and activities ........................................................... 52
Figure 4.1.     Service level of knowledge markets ........................................................ 63
Figure 4.2.     Knowledge-market dimensions ............................................................... 66
Figure 6.1.     Venture capital is highly cyclical .......................................................... 101
Figure 6.2.     Shift in balance between investment in tangible/intangible assets in
                the United States.................................................................................... 107


Table 3.1.      Consortia focus by disease area............................................................... 49
Table 3.2.      Policy objectives for KNMs .................................................................... 58
Table 4.1.      Elements for the representation of knowledge goods in a digital
                knowledge object..................................................................................... 64
Table 5.1.      GAIN data-access and data-access policies ............................................ 82

                                                                        EXECUTIVE SUMMARY –   9

                                      Executive summary

            There is a proliferation of initiatives in the life sciences to bring together
       dispersed and diverse elements of the research infrastructure and simplify
       the process for learning about, accessing and utilising sometimes dispersed
       knowledge and intellectual assets. The common goal of these initiatives is to
       leverage innovative capacity by creating interconnected webs of knowledge
       that exploit external expertise. Recent advances in information technology
       make these initiatives possible: the storage capacity of data; its ease of
       transmissibility across the Internet; the development of software to access,
       make interoperable and analyse data, as has the creation of governance
       systems that regulate access and use of data. In this report, we refer to such
       initiatives as “knowledge networks and markets” (KNMs).
           This report describes a range of different types of initiatives, in
       particular: i) data registries and repositories; ii) platform technology and tool
       providers; iii) research consortia and public private partnerships; iv) intel-
       lectual property pools, clearinghouses and exchanges; and v) prizes, online
       auctions, brokers and citizen science projects. Most of the extant initiatives
       are best described as networks, where an effort is made to improve the
       pooling, access to and exchange of knowledge. But some early experiments
       in markets are emerging where the goal is the easier monetisation and trade
       of knowledge in the form of intellectual assets.
            While there are many dozens of active knowledge networks and markets
       in the life sciences, their results and impact on the pace and direction of the
       innovation process have yet to be well understood. For some, these arrange-
       ments are neither new nor disruptive. For others, however, the proliferation
       of initiatives is a sign that a new knowledge infrastructure is emerging that
       will reduce knowledge search and transaction costs, create value from
       underused assets, and broaden the innovator base to new potential “problem
       solvers” and innovators. There are no agreed metrics of their success. And
       indeed, the approaches taken to knowledge generation, management and
       exchange are far from uniform, making agreement on what to measure
       difficult. KNMs differ in terms of what knowledge type is being collected
       and exchanged, the structure of the organisations created, and their financing


           The report teases out what are the common features of KNMs and on
       what dimensions they vary. In particular it concludes that most KNMs are
       formal, which means that their creators need to address governance issues
       and IP rights; they are transformative of the knowledge collective because
       they reorganise resources and facilitate inter-organisational relationships;
       and they are translational because they facilitate the further development of
       data, information and knowledge.
           KNMs are of enormous interest in industry, in government agencies and
       in the not-for-profit research organisations. All of these types of
       organisations, each in their own way a key contributor to the innovation
       process in the health sector, are actively participating in the creation of new
       KNMs as they each search for new business models that can better deliver
       the next generation of health technologies. KNMs are especially prized as a
       means for reducing the costs and risks of precompetitive research, and even
       as a way of broadening what is collectively defined as “pre-competitive”.
       KNMs are also deemed vital in the era of big data because they allow
       experimentation with new approaches to the collection, integration and
       analysis of biomedical data. Finally, some KNMs seek the involvement of
       regulatory agencies as they modernize drug development and offer a more
       open dialogue between innovators and regulators about how to evaluate new
       biomedical technologies and how to collaboratively develop common
       standards and platform technologies.
            For policy makers, the rise of KNMs raises several questions. First, why
       are new organisational models for R&D in the life sciences emerging?
       Second, do KNMs represent a break with past models of innovation systems
       in the life sciences? Third, what is the impact of such distributed, networked
       systems of R&D on the productivity of health innovation? Fourth, how do
       government actions encourage or discourage the formation and effectiveness
       of knowledge networks and markets, and should governments be more
          The main messages that emerge from the analysis in this report are
       summarised as follows:
           1. There is a shifting culture and new acceptance of more open
              innovation strategies, and as a result new business models and
              organisations are emerging to take advantage of the greater reliance
              on outsourcing R&D.
           2. A more networked and interoperable research infrastructure has
              enormous scientific and commercial potential. The use of informa-
              tion technologies (e.g. data collection and coding, storage and
              analysis) is enabling novel approaches to research, the crossing of
              disciplinary boundaries and the exploitation of efficiencies.

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                    EXECUTIVE SUMMARY –   11

            3. New business models and ways of organising R&D are emerging as
               improvements are made in our ability to represent knowledge
               objects in electronic marketplaces and in the valuation of knowledge
               intensive intellectual assets. It is too early to tell which are
               sustainable or whether there are emerging best practices. However,
               the proper valuation of intellectual assets will be at the heart of
               effective exchange and trading mechanisms, which may themselves
               create demand pull for better husbanding of knowledge and an
               acceleration of effort may be desirable here.
            4. Improving the interoperability of knowledge resources is funda-
               mental in creating the infrastructure that allows KNMs to emerge.
               But it is a major undertaking and has important technical, semantic
               and legal components that must be tackled together. In creating new
               bioinformatics technology infrastructures one should strive to be
               “technology neutral” so that systems are adaptable and do not limit
               the future scope of research or collaborations.
            5. Long-term sustainable funding is necessary to establish and
               maintain KNMs and should include human-capital investments.
               This may have implications for how biomedical research is funded
               in the future if greater importance is placed on promoting pre-
               competitive research.
            6. Government policy will affect KNMs through mechanisms such as
               R&D funding, data access and sharing policies, intellectual property
               right (IPR) policies, grant stipulations, infrastructure and project
               funding, the creation of consortia or PPP, competition policy, and
               privacy and security policies.
            7. There is a need clearly to identify the incentives driving the creation
               of and participation in KNMs and to monitor their effectiveness in
               order to understand what makes such organisations and/or busi-
               nesses successful and sustainable over the long term and for what
               sorts of scientific endeavours.
            8. In areas where there is a strong public policy interest, governments
               can play a catalytic role in bringing diverse parties to the table to
               discuss new knowledge exchange and generation mechanisms. But
               as governments become increasingly involved in KNMs – as
               developers, funders or partners – safeguards related to access and
               equity will need to co-evolve.


           OECD countries have a strong policy interest in knowledge markets as a
       tool to achieve many economic and social goals in health and even more
       broadly in the life sciences. There are two specific areas which drive current
       government interest in KNMs:
           •   Large-scale shared infrastructures platforms for life sciences. These
               large-scale infrastructures will enable the creation of new tools in
               biotechnology, particularly with the convergence of new biology,
               IT, physical sciences. Governments would like to use lessons from
               the current generation KNMs as they develop policy principles for
               integrating complex, high-value data and new IT platforms.
           •   Social networking in the life sciences. Substantive international
               efforts are being made to enable social networking in life sciences.
               However, these efforts should be driven by a clear policy commit-
               ment and broadly accepted operational guidelines.
            Given the relative novelty of KNMs, there needs to be on-going
       monitoring of performance and outcomes that test the assumptions on which
       KNMs are being built. Some elements of common infrastructure – and
       policy assumptions – will be necessary for KNMs to function as desired, but
       the jury remains out on many of the specifics of such. For example,
       monetised knowledge markets might help spread risk and bring otherwise
       underutilised inventions through to the market – but better means to
       evaluate, report and trade intellectual assets need to be developed to make
       for real dynamism.

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

                                             Chapter 1

            The rise of knowledge networks and markets as
                      enablers of open innovation

      Within the past five years there has been an explosion in the development of
      knowledge networks and markets (KNMs) in the life sciences. This chapter
      sets out the context for the development of these phenomena, provides a
      rudimentary operational taxonomy, and sets out some overarching questions
      about their impact on innovation productivity as well as what governments
      might need to do to influence their success.


          Most observers of the life sciences and the biopharmaceutical industry
      agree that an organisational change in the innovation process is afoot. The
      digitisation of biomedical information; the increased capacity to store,
      integrate and evaluate data; and the advent of the Internet and the World
      Wide Web have created unprecedented opportunities to improve information
      flow and exchange in the life sciences. Financial pressures on the pharma-
      ceutical industry and changing public policy priorities in health care also
      have created an opportunity for the emergence of new collaborative arrange-
      ments and institutions in pharmaceutical and biotechnology business prac-
      tices. Much of the pharmaceutical industry now claims to operate within this
      business space of much more open, partnership-based innovation (see, for
      example, recent annual reports of GlaxoSmithKline and Pfizer [GSK, 2010;
      Pfizer, 2010]).
          This report introduces the concept of knowledge networks and markets
      (KNMs) as a vital element of this more open innovation model and argues
      that the creation of these new infrastructures is an important trend in the life
      sciences, particularly in the health sector. Previous work at the OECD has
      identified barriers to increased health innovation productivity due to the
      prevailing practices of keeping knowledge assets secret or proprietary. New
      knowledge access and collaboration tools hold promise for accelerating the
      bench-to-bedside translation of basic science into commercialisable health
      innovations that meet social needs. Networked approaches to research and
      development (R&D) can move biomedicine toward a more systemic, holistic,
      evidence-based understanding of health and disease.
          Few fully understand the nature of the new exchange mechanisms, the
      forces that are driving their creation, the extent of their use, or their implica-
      tions for the innovation process. Is the trend toward a more open approach a
      radical change that will alter what we understand science to be and speed the
      rate of discovery? Or is it an adaptation to the realities of information
      technology and the Internet that will not significantly change the culture of
      the biomedical scientific enterprise? Policy makers, in particular, need more
      information about the role that KNMs can play in the emerging bioeconomy
      as well as guidance on the role policy plays in facilitating their emergence
      and effectiveness.
          In this chapter we define what is meant by KNMs and provide examples
      of a few different types of these arrangements in order to familiarise the
      reader with the range of networks and markets that are emerging to
      exchange and trade knowledge intensive assets in the life sciences.

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

Definition of knowledge networks and markets (KNMs)
            KNMs include a broad range of existing initiatives whose purpose is to
       improve access to widely distributed biomedical knowledge resources in order
       to facilitate further innovation. The goal is to improve the circulation (sharing,
       trading or joint production) of disembodied knowledge (inventions, IPR,
       software, data, know-how) between independent parties using various vehicles
       (commercial transactions like licensing, spillovers, joint facilities, special
       interfaces, individual mobility, mergers and acquisitions, direct investment).
       The purpose of a knowledge network or market thus is to foster the more
       efficient use of knowledge and to enable cumulative innovation.
           There is no commonly accepted terminology to capture the range of on-
       going experiments in how organisations in the life sciences access and use
       knowledge resources. Perhaps the term that is most familiar is “open” as in:
       open source, open access, open science and open innovation. Initiatives that
       are billed as open, however, vary substantially in terms of what exactly is to
       be kept open and what “open” actually entails. “Open”, over the last twenty
       years, has come to signify a number of different things:
            •    “Open source” referred originally to the development and
                 improvement of software (source code in particular) by multiple
                 different users working independently and remotely. The Linux
                 operating system was developed in the early 1990s using Linus
                 Thorvald’s new development approach, which can be described as:
                 “release early and often, delegate everything you can, be open to the
                 point of promiscuity” (Raymond, 2001). Open source is now applied
                 more broadly to designate innovations, often jointly developed by
                 different contributors, available royalty free to anyone and without
                 restrictions on how they are to be used.
            •    The “open science” movement wants to foster greater transparency
                 in the scientific methodology used and data collected, and to ensure
                 the public availability and reusability of data, tools and materials.
                 Open science argues for broadly and openly communicating research
                 and its results.1 “Open science is the timely communication of
                 research, from start to finish, without boundaries, to enable a billion
                 minds to participate.”2
            •    “Open access” is about making scientific literature and data “digital,
                 online, free of charge, and free of most copyright and licensing
                 restrictions.”3 In open access, innovators purposely do not seek IP
                 rights. Some open access successes include the creation of line
                 repositories for journals, new types of free but peer reviewed on-line
                 journals, and new access policies regarding publications and research
                 results that apply to academic faculty or recipients of public funds.


          In all the above cases, Internet-based tools have been critical in
      improving transparency and access to the technical and scientific knowledge
      base.4 The hope is that earlier communication of problems, methodologies,
      data, tools and results, will allow mass collaboration on or even crowd
      sourcing of scientific and technical problems, and ultimately the more rapid
      generation of new knowledge and the discovery of better solutions.5
          There is another type of “open” which is relevant to KNMs, but which
      applies explicitly, though not exclusively, to the private sector.
          •   “Open innovation” is “the use of purposive inflows and outflows
              of knowledge to accelerate internal innovation, and expand the
              markets for external use of innovation.”6 Firms that adopt open
              innovation business models make active use of licensing, colla-
              borations, joint ventures, spin-offs or acquisition of firms. They do
              not rely exclusively on their in-house research or development
              capacities and believe that unused ideas should be offered to outside
              firms for further development and commercialisation (OECD,
              2008). Open innovation means that a firm’s intellectual assets are
              actively managed and the paths to commercialisation diversified.
              Open innovation recognises that “the boundaries between a firm and
              its environment” are permeable (Chesbrough, 2003).
          Thus there are several, sometimes incompatible, meanings of “open”,
      including: creating a more accessible pre-competitive science base; enabling
      multiple independent innovators to work on the same problem; and
      knowledge management where innovations are strategically transferred in
      and out of the firm. Openness always strives to either improve or accelerate
      the innovation process, but the means by which this openness is accom-
      plished and the incentives driving it differ significantly.
           In this report we focus on the emergence of a number of institutions that
      strive to make the life sciences more open. We use the term “knowledge
      networks and markets” in order to better capture the characteristics of these
      institutions and arrangements. KNMs all facilitate the search for knowledge and
      the matching of knowledge resources to those who can use them. Networks
      refer to stable connections (links) between separate entities (nodes). Knowledge
      networks allow the more continuous flow of knowledge between innovators and
      potential users. Examples of networks are consortia, public private partnerships,
      and data repositories. Typically, networks do not involve monetary transactions
      for knowledge exchange, and they often encourage collaboration amongst
      entities. Knowledge markets include arrangements that permit more discrete
      transactions for knowledge resources (for example through auctions, broker-
      ages, pools, clearinghouses) often as commercial exchange. Markets sometimes
      price the knowledge resources being exchanged.

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

            KNMs are understood to be:
            •    The variety of institutional arrangements that govern the circulation
                 (sharing, trading or joint production) of disembodied knowledge
                 (inventions, IPR, software, data, know-how) between independent
                 parties. KNMs allow transactions to take place in a context of trust
                 thus reducing transactions costs.
            •    KNMs are platforms that facilitate the search for and use of
                 knowledge. They increase the range of potential users of knowledge
                 assets or research partners often in areas where there are collective
                 action problems.
            •    KNMs massively leverage innovative capacity by creating inter-
                 connected webs of knowledge, exploiting external expertise (from
                 outside organisations, disciplines, national borders), and sometimes
                 enabling crowd sourcing.
            •    KNMs foster the efficient use of knowledge and enable cumulative
                 innovation, with the aim of developing new knowledge or producing
                 goods and services.

Varieties of KNMs

           KNMs are relatively new in the biological sciences, when compared to
       computer science, which spawned the open source movement, and physics,
       which created the Internet and where large scale collaboration and big data
       sets have a longer history. Perhaps one of the first knowledge networks in
       biology was the Human Genome Project (HGP). Launched in 1990, the US
       Human Genome Project set out to create a public database of the sequence
       of the three billion base pairs in the human genome and a map for the
       location of all human genes. It was mostly funded by the United States
       government (through the Department of Energy and the National Institutes
       of Health) and at the outset was construed as a top-down big science project
       with three large-scale sequencing centres funded for the explicit purpose of
       carrying out the HGP. There were five main institutions at the launching of
       the consortium, but by 2003 there were over twenty different laboratories
       from six countries participating. The work of this international consortium
       was co-ordinated through meetings and calls, and staff was recruited to
       perform specific tasks from sequencing technology to computer analysis
       (Collins et al., 2003). In short, the HGP was not an “open source” project in
       that there was an agreed strategic plan and timeline for its accomplishment.
       But the objective of the HGP was decidedly about “open access”: all colla-
       borating laboratories agreed to submit sequence data to a common publicly
       available database that was published, ahead of plan, in 2003.


          Since, the number of KNMs in the life sciences have proliferated,
      including the International HapMap Project to identify and catalogue genetic
      similarities and differences in human beings in 2003; the International
      Cancer Genome Consortium to characterise genetic mutations that are
      involved in a variety of tumours (Anonymous, 2010); and the Human
      Microbiome Project (HMP) to characterise the microbial communities found
      at several different sites on the human body so as to understand their role in
      health and disease.
           Genetics was at the forefront of the open science movement, perhaps
      because nucleotide sequences were easy to characterise and collect digitally.
      But there are now dozens, if not hundreds, of initiatives in the life sciences
      which bring together dispersed and diverse scientific contributors (from
      different disciplines and different types of research centres) and whose aim
      is to simplify the process for learning about and accessing the knowledge of
      the broader research community in order to accelerate innovation. Some
      espouse open source or open science goals, others focus on open access, and
      still others are interested in open innovation or networked approaches to
      science.7 In Chapter 5 we describe a number of these KNMs.
          However, it is worth noting from the outset that within the broad
      umbrella of KNMs there are many varieties of initiatives and many types of
      participating organisations. More in depth descriptions can be found in the
      recent OECD report on “Collaborative Mechanisms for Intellectual Property
      Management in the Life Sciences” (OECD, 2011).

      Data registries and repositories
          These are online infrastructures that allow institutions or individuals to
      access, use and/or contribute data and information that is of biological or
      medical relevance. One of the oldest such repository is probably GenBank,
      which was founded in 1982 by the National Institutes of Health (NIH) and
      makes available an annotated collection of all publicly available DNA
      sequences. In the United States researchers who received federal funds and
      many journals require that genetic data in articles be deposited in GenBank
      prior to publication.
          A more recent database is the NIH funded The Cancer Genome Atlas.
      This large-scale, high-throughput effort is being carried out by a network of
      more than 100 researchers who collect cancer tumour samples, and then
      sequence and analyse the cancer cells in order to systematically catalogue
      the genomic changes that occur in more than 20 common types of cancer.
      According to their website: “The TCGA Research Network is actively
      depositing comprehensive data sets into publicly accessible databases. Any

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

       researcher can leverage these data to generate and/or test hypotheses,
       validate their own work or add power to another data set.”8
           Other data repositories include things like the newly created Virus
       Pathogen Resource Bioinformatics Resource Center, which is an open-
       access online database and analysis resource centre to help scientists study
       and combat human pathogenic viruses. ViPR provides genomic and pro-
       teomic data, information about the characteristics of viruses, information
       about the immune response to viral infections, human clinical data and
       surveillance data.
           A radically different model of a data registry is Patients Like Me. is an on-line social networking site that allows
       patients to exchange information about their disease and its treatment.
       Patients voluntarily submit data on their health status, symptoms and moods,
       treatments and side effects. The information is open, so that it can be shared
       and compared with other patients. But Patients Like Me also has an internal
       research team and collaborates with academic projects. It also makes de-
       identified data available for purchase to companies that will use it to
       improve or understand products or markets.

       Platform technologies and tools
           Some projects and institutions go beyond making data accessible and
       offer curated or linked databases or tools as publicly available resources.
       Tools include things like model chemical or biological materials, model
       organisms or virtual models, and analytical tools. The costs of building,
       maintaining and offering some of these services can be substantial so while
       some of the platform technologies and tools are “free”, many others are fee-
           For example, a not-for-profit group, Sage Bionetworks is building
       complex, predictive models of disease for use by researchers. Sage
       Bionetworks facilitates the integration of the vast amounts of large-scale
       biological information that is being generated — for example, from imaging
       studies, microarray analysis and next-generation sequencing — into
       models.9 It hopes to convince biologists to pool raw experimental data into
       the public domain so that models can be built. Sage Bionetworks provides
       open access to selected datasets and network models, and explicitly
       forswears intellectual property rights in the data. Nevertheless it is also
       working in more traditional fee for service collaborations with individual


          Another not-for-profit, the BioBricks Foundation, is creating a registry
      and repository of standard biological parts that are the building blocks of
      synthetic biology. A “BioBrick™ standard biological part is a nucleic acid-
      encoded molecular biological function […] along with the associated
      information defining and describing the part.”10 Scientists can browse the
      Biobricks catalogue and contribute new ones that conform to the Founda-
      tion’s specification. The BioBrick foundation also provides a model contract
      “that allows individuals, companies, and institutions to make their stan-
      dardized biological parts free for others to use.”11 BioBricks has created a
      technical standard, an open technology platform, and a repository open to
      anyone interested in building new biological parts.
           The private sector is also exploring making some of its platform
      technologies more accessible, often for particular causes. In 2010
      GlaxoSmithKline made publicly available data and compounds believed
      relevant for the development of new medicines for malaria. GSK screened
      two million molecules in its compound library for reactions to the malaria
      parasite P. falciparum and found more than 13 500 compounds that
      inhibited the parasite. GSK committed to making the compounds, their
      chemical structures and associated assay data freely available to the public
      on leading scientific websites in order to encourage further research by the
      scientific community on the compounds and bring more minds to bear on
      this challenging problem.

      Consortia and public-private partnerships
          Consortia and public-private partnerships are formal agreements
      between parties engaged in a common project, in which the contributions to
      be made by different parties and the resources to be pooled are defined, as
      are the terms under which outcomes from the collaboration will be shared or
      made accessible to others. By definition, consortia and public-private
      partnerships have limited membership, they are not open to any and all. In
      some cases, they are committed to making the fruits of their collaboration
      publicly accessible.
          In Europe, the Innovative Medicines Initiative (IMI) is a public-
      private partnership (PPP) between the pharmaceutical industry and the
      European Union (represented by the European Commission) to address drug
      R&D bottlenecks. Its objective is to improve the ability to predict the safety
      and efficacy of an investigational compound as early as possible in drug
      development through improved knowledge sharing and management. In
      individual projects, collaborating partners — including companies, public
      research organisations and regulators — must share resources and expertise
      (for example, clinical data and samples from past development programmes)
      to address a particular challenge in drug discovery and development. IMI

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

       explicitly sets out both to create new and more effective networks between
       pharmaceutical companies and their governmental and research organisation
       partners, and to mobilise knowledge and share information within the
       members of the PPP that would otherwise be kept secret by firms.
            The Biomarkers Consortium is a platform for pre-competitive col-
       laboration designed to accelerate the development of recognised new
       biomarkers in specific therapeutic areas such as cancer, neuroscience and
       metabolic disorder. In the consortium, companies collaborate with each
       other and with the US Food and Drug Administration (FDA) to evaluate the
       utility of proposed biomarkers. Within each therapeutic area project,
       companies pool relevant data from research and clinical trials they have
       conducted, jointly analyse it, and come to an agreement about what
       conclusions can be drawn about a specific biomarker. The results of their
       deliberations are openly published, so that other researchers or firms can use
       the biomarker in their research and development activities. What is new here
       is the sharing of data and the agreement amongst a network of companies
       that biomarker utility evaluation is actually pre-competitive. Companies
       choose to participate because they can advance the use of biomarkers, both
       for their own research and in the regulatory process.

       Pools, clearinghouses and exchanges
           Another approach to increasing access to particular knowledge assets is
       through the creation of formal agreements between parties – such as patent
       pools, clearinghouses, or exchanges. In these institutions, parties agree to the
       terms by which intellectual assets are submitted for sharing and the terms on
       which they can be used by either the participants in the group or by the
       broader community. Unlike consortia or public private partnerships, there is
       no common project or formal research collaboration. The idea is simply to
       create a common resource, for convenience, to reduce transaction costs in
       negotiating access to intellectual assets, and to create a trusted intermediary
       for doing so. Pools, clearinghouses and exchanges do provide transparency
       as to what intellectual assets they hold. The extent to which access to those
       assets is available to everyone and the cost of access can vary.
           One patent pool explicitly designed to encourage mass collaboration in
       the development of medicines for neglected tropical diseases was started by
       GlaxoSmithKline in 2009. When first created, the pool comprised only GSK
       intellectual property (IP), containing over 800 granted or pending patent
       applications accessible to researchers. However, the goal from the outset
       was that other companies should also put their IP in the pool. Alnylam
       Pharmaceuticals was the first company to join, donating over 1 500 patents
       (issued or pending) on its RNA interference technology. In addition to its IP,
       Alnylam will also provide technical know-how on a royalty-free, non-profit


      basis in the least-developed countries through licensing agreements with
      qualified third parties. It is hoped that ultimately the pool will include
      patents from numerous pharmaceutical companies, biotech firms, patient
      groups, NGOs and universities. Groups donate to the pool relevant small
      molecule compounds or process patents for neglected tropical diseases, and
      allow others access to develop and produce new therapies and formulations
      for use in the least-developed countries. Two agreements for access to the
      pooled patents have recently been signed, and the South African government
      has announced its intention to use the pool.

      Prizes, on-line auctions, brokers and citizen science projects
          Finally some open science initiatives are designed to facilitate “fishing
      expeditions” for relevant knowledge. Instead of unilaterally building a
      common resource for the use of all, these initiatives try to attract new talent
      to tackle discrete problems that innovators need solved. By putting their
      technical or scientific challenge out in the open, individuals, companies or
      governments attract external talent to their cause. Prizes award money to the
      individual or group that best meets the criteria set out in the prize
      announcements; on-line auctions and brokerages pay innovators that
      successfully solve an advertised problem; and in citizen science projects
      innovators, formally qualified or not, participate in initiatives for fun with
      no remuneration at all. The extent to which the results of these scientific
      transactions are made public depends very much on the goals of the ones
      posting the challenges. These initiatives openly advertise their problem and
      goal, create a global network of innovators thus enabling collective
      innovation, but may or may not decide to make solutions publicly available.
          •   Innocentive is an on line bulletin board of technology where
              seekers of solutions challenges in all fields, including health
              sciences, are brought together in such a way that allows companies
              and research organisations to farm out research questions for which
              they do not have in-house capabilities in exchange for a stated
              award. Researchers can choose to submit solutions to the challenge
              and the winner of the challenge if there is one, collects the award. It
              is a model of “hive” pharmaceutical development.
          •   Prize4Life is a not-for-profit entity which focuses on accelerating
              the development of new therapies for ALS (amyotrophic lateral
              sclerosis or Lou Gehrig's disease). Prize4Life offers USD 1 million
              prizes for biomarkers that help track the progress of the disease and
              for treatments that are proven to be effective in mouse models.
              Prize4Life tries to attract new researchers and new ideas to the ALS
              field. By identifying major gaps in the drug development process,
              they help leverage existing projects and expertise.

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

            •    Foldit is an online game which enlists players worldwide to solve
                 difficult protein-structure prediction problems. It uses humans'
                 pattern-recognition and puzzle-solving abilities to compete against
                 existing computer programmes at pattern-folding tasks (Cooper
                 et al., 2010). Foldit players – most of whom have little or no
                 background in biochemistry – are able to solve protein structure
                 refinement problems for proteins implicated in human health and
                 disease (Khatib et al., 2011).

A diversity of structures

           The examples above illustrate the variety of recent initiatives and are not
       meant as an exhaustive list. Nor are the categories identified mutually
       exclusive. Some initiatives fall under multiple categories because they offer
       a variety of different services.
           KNMs clearly are not a single institutional arrangement. They comprise
       multiple organisational forms that advance open access, open source, open
       science or open innovation objectives. The different forms of KNMs can
       vary in a number of important respects:
            •    KNMs operate in different fields with different technologies at
                 different stages of development. Some initiatives focus simply on
                 data collection, others on building models or platforms, others on
                 solving scientific or technical problems.
            •    KNMs differ in their size, level of funding and ambition. Individual
                 scientists experiment with “open notebooks”. The US National
                 Cancer Institute is trying to make all cancer related data from
                 scientific laboratories to clinics interoperable. Other KNMs are
                 international consortia requiring financial commitments from multiple
            •    KNMs can facilitate one-off exchanges of discrete information; they
                 can also build longer term networks of collaboration amongst parti-
                 cipants; or they can encourage crowd sourcing and a diversity of
                 approaches to the use and analysis of scientific resources.
            •    Interactions between participants in KNMs can be entirely on-line
                 and virtual; other KNMs require face-to-face meetings for discus-
                 sions or negotiations, about for example the science or functioning
                 of the institution.
            •    The parties involved in KNMs can include both for-profit and not-
                 for-profit entities, government agencies, and even individuals.


          •   Some of the knowledge transactions are commercial and others are
          •   KNMs impose different requirements on participants depending on
              the type of KNMs involved and the type of user: financial invest-
              ments; commitments to the joining of a consortia; the contractual
              commitment to provide certain type of data and knowledge to
              partners or publicly.
          •   Finally, the initiatives vary in their “openness”: both in how open
              they are to participation by the broader research community or
              public and also in how “open” the results of their collaboration are
          We are in the midst of an explosion of experiments that each take
      advantage of advances in IT which vastly expand our ability to collect,
      analyse and distribute knowledge. What all knowledge networks and
      markets share is the goal of advancing science and developing new
      knowledge or producing new goods and services. All the institutions that
      will be discussed in this report are deeply interested in creating a new
      culture of co-operative, data-intensive science.

Open questions about KNMs

          Fundamentally, policy makers are interested in whether these initiatives
      are indeed the basis for a new distributed, networked system of R&D that
      significantly improves the productivity of health innovation. Given the
      diversity of experiments in open science, policy makers do not yet
      understand how KNMs will reshape science, technology and innovation
      policies. Governments are already investing in the creation of KNMs, and
      thus have played a role in defining their governance. But there remain many
      questions about the incentives to participate, the design of these institutions,
      their sustainability, and their evaluation.
          The OECD Working Party on Biotechnology (WPB) has concluded that
      the scientific and economic impacts of KNMs are likely to be important:
          •   There is a shifting culture and new acceptance of more open
              innovation strategies, and as a result new business models and
              organisations are emerging to take advantage of the greater reliance
              on outsourcing R&D.
          •   New business models and ways of organising R&D are emerging as
              improvements are made in our ability to represent knowledge
              objects in electronic marketplaces and in the valuation of knowledge

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

                 intensive intellectual assets. It is too early to tell which are sustain-
                 able or whether there are emerging best practices.
            •    A more networked and interoperable research infrastructure has
                 enormous scientific and commercial potential. The use of informa-
                 tion technologies (e.g. data collection and coding, storage and
                 analysis) is enabling novel approaches to research, the crossing of
                 disciplinary boundaries and the exploitation of efficiencies.
           The WPB also conclude that before the potential of KNMs can be fully
       realized there are some obstacles to their widespread use as either scientific
       or business platforms. In particular:
            •    There is a need clearly to identify the incentives driving the creation
                 of and participation in KNMs.
            •    Long-term sustainable funding is necessary to establish and main-
                 tain KNMs.
            •    The proper valuation of intellectual assets will be at the heart of
                 effective exchange and trading mechanisms, which may themselves
                 create demand-pull for better husbanding of knowledge.
            •    KNMs should be monitored for their effectiveness, to understand
                 what makes such organisations successful and sustainable over the
                 long term and for what sorts of scientific endeavours.
           The role of government in establishing or supporting KNMs is impor-
       tant. Various existing policies have an impact on the ability to form KNMs
       and some new policy measures seem necessary. For example:
            •    Government policy will affect KNMs through mechanisms such as
                 R&D funding, data access and sharing policies, intellectual property
                 right (IPR) policies, grant stipulations, infrastructure and project
                 funding, the creation of consortia or PPP, competition policy, and
                 privacy and security policies.
            •    In areas where there is a strong public policy interest, governments
                 can play a catalytic role in bringing diverse parties to the table to
                 discuss new knowledge exchange and generation mechanisms. But
                 as governments become increasingly involved in KNMs – as
                 developers, funders or partners – concerns about access and equity
                 may arise.
            •    Government funding for the launch or maintenance of certain
                 KNMs will be necessary and will include human-capital invest-
                 ments. This may have implications for how biomedical research is
                 funded in the future if greater importance is placed on promoting
                 pre-competitive research.


          •   Improving the interoperability of knowledge resources is funda-
              mental in creating the infrastructure that allows KNMs to emerge.
              But it is a major undertaking and has important technical, semantic
              and legal components that must be tackled together. In creating new
              bioinformatics technology infrastructures one should strive to be
              “technology neutral” so that systems are adaptable and do not limit
              the future scope of research or collaborations.

Report structure

          In this report we explore: i) What characterises KNMs and how are they
      different from older forms of co-operation and exchange? ii) What are their
      likely scientific or economic impacts and can these be measured? iii) What
      policies help or hinder the development of KNMs?
          The following chapters will take up these themes in greater detail. Chapter 2
      reviews the dynamics of the innovation process and identifies some of the
      inefficiencies in the current biomedical research system. It sets the stage for
      understanding why there is so much interest in sharing biomedical knowledge
      more widely to create added value.
           Chapter 3 explores how different sectors and groups perceive the potential
      advantages that KNMs present in comparison to more traditional strategies for
      ensuring flows of knowledge between organisations. It identifies why firms
      see KNMs as an important addition to their existing innovation strategies and
      why governments have an interest in supporting their development. It also
      addresses what policies are being used to foster the creation of KNMs, as well
      as factors that might limit their development.
           Chapter 4 presents different theoretical approaches to the design of
      KNMs. The chapter draws on insights about knowledge management from the
      field of computer science. The chapter discusses the difficulty and complexity
      inherent in creating a useful, interoperable infrastructure of biomedical data. It
      argues for creating resources that are not tied to particularly technologies and
      can be nimbly adapted as users’ needs evolve.
           Chapter 5 presents case studies of currently functional KNMs in the life
      sciences. The case studies include information about the purpose of the
      initiative, its organisation and structure, its participants and the sorts of
      knowledge resources that are exchanged.
          Chapter 6 discusses how knowledge in knowledge markets is valued and
      the importance of valuation for scaling up knowledge markets. It also
      identifies some of the difficulties in expanding the use of knowledge
      markets in the present environment.

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

            Chapter 7 presents conclusions and policy recommendations. It identifies
       what are the common features of KNMs and on what criteria they differ. It
       reiterates why KNMs matter for the innovators in the biopharmaceutical
       sector, in particular because of their contribution to: addressing financial
       pressures, accelerating science and development, improving health outcomes,
       and improving the regulatory dialogue. The chapter concludes with a review
       of the support that KNMs have from government in their establishment and
       maintenance stressing the role of ICT infrastructures, intellectual property
       rights, intellectual asset valuation, and regulatory frameworks. Future work
       at the OECD and elsewhere could focus on the role that KNMs play in
       enabling social networking in the life sciences and in transforming large-
       scale shared infrastructure platforms.



1.    See, for example, Science Commons’ Principles for Open Science at
      science/. OpenScience Blog,, accessed on
      29 November 2011.
2.    From Cameron Neylon’s blog, Science in the Open, at, accessed on 30 November 2011.
3.    From Peter Suber’s Open Access Overview website at, accessed on 30 November 2011.
4.    These projects use online tools as cognitive tools to amplify our collective
      intelligence. The tools are a way of connecting the right people to the right
      problems at the right time, activating what would otherwise be latent expertise.
5.    For a discussion of crowd sourcing research see Gower, “Is Massively
      Collaborative Mathematics Possible?” at
      mathematics-possible/, accessed on 29 November 2011.
6.    From the University of California at Berkeley’s Open Innovation Program,, last accessed on
      5 December 2011.
7.    The variety of goals and models on Knowledge Markets and Networks in the
      life sciences seem to contradict the thesis of some authors that “open source” is
      emerging as a solution to some of the innovation challenges in this field. See
      for example, Hope (2008).
8.    See The Cancer Genome Atlas website:
9.    See
10.   See, accessed on 7 December 2011.
11.   ibid.

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

Anonymous (2010), “International Network of Cancer Genome Projects,” Nature
     464, 15 April, pp. 993-998.
Chesbrough, H. W. (2003), “The Era of Open Innovation,” MIT Sloan
      Management Review, Spring 2003, Vol. 44, Issue 3, pp. 35-41.
Collins, F. M. Morgan and A. Patrinos (2003), “The Human Genome Project:
      Lessons from Large-Scale Biology,” Science, 11 April, pp. 286-290.
Cooper, S., F. Khatib, A. Treuille, J. Barbero, J. Lee, M. Beenen, A. Leaver-Fay,
     D. Baker, Z. Popovi (2010), “Predicting protein structures with a
     multiplayer online game.” Nature 466, 5 August, pp. 756-760.
     See also:
GSK (GlaxoSmithKline) Annual Report (2010),
Hope, J. (2008), Biobazaar: The Open Source Revolution and Biotechnology,
      Harvard University Press: Cambridge.
Khatib, F., et al. (2011), “Crystal Structure of a Monomeric Retroviral Protease
      Solved by Protein Folding Game Players,” Nature Structural & Molecular
      Biology, 18, 18 September, pp. 1175-1177.
OECD (2008), Open Innovation in Global Networks, OECD Publishing, Paris.
OECD (2011). “Collaborative Mechanisms for Intellectual Property Management
    in the Life Sciences”, OECD, Paris.
Pfizer Annual Review (2010),
Raymond, E., (2001), “The Cathedral and the Bazaar”.,
     Last accessed on 1 December 2011.

                                                                    2. KNOWLEDGE FLOWS –   31

                                             Chapter 2

                                       Knowledge flows

      Knowledge is at the core of innovation and needs to flow effectively if
      productivity is to be efficient. But knowledge also has great value and is
      jealously guarded by innovators. This chapter explores how the “knowledge
      complex” – the flow and exchange of knowledge – in the biosciences is
      changing as the volume of knowledge increases and the numbers and variety
      of actors involved in knowledge creation and use grow. The functioning of
      knowledge networks and markets are set in the context of this new knowledge


         “Knowledge has become the most important factor in economic life […]
         Intellectual capital […] has become the one indispensable asset of
         corporations.”                                       (Stewart, 1997)

          Ever since Peter Drucker popularised the concept of a knowledge
     economy in the late 1960s (Drucker, 1969) policy makers, business leaders
     and academics have striven to capture the economic and political power
     associated with the creation, development and use of knowledge. A signifi-
     cant literature on the management of knowledge took off from the mid-
     1990s (Gordon and Grant, 2000) and intellectual assets became spoken of as
     a “new” currency of commerce and of growth. By the second decade of the
     21st century, these efforts to grow economies based around knowledge have
     if anything intensified.
         In fact, the realisation that knowledge is an important economic and
     political commodity is hardly new. The great 19th century economist, Alfred
     Marshall, for instance, wrote “Capital consists in a great part of knowledge
     and organisation […] Knowledge is our most powerful engine of production”
     (Marshall, 1972). But the costs and efforts associated with generating and
     accessing knowledge – let alone applying it – have changed very significantly
     over the last century, thus the increasing focus on valuing knowledge
     (intellectual) assets and managing knowledge.
         The main factors driving these changes include: the acceleration of
     change in markets, competition and technology; the globalisation of the
     production of new knowledge, with new actors, like China, becoming
     increasingly significant (see Figure 2.1); and the ability to communicate and
     exchange data and information incredibly quickly through the Internet.
         These developments have brought about a new generation of knowledge
     interdependence (Quintas et al., 1997), where no one firm or actor can hope
     to deliver all the knowledge necessary to put together competitive, state of
     the art technological innovations. This increasing knowledge interdependence
     has brought with it a dynamic of knowledge flow, whereby knowledge – and
     the technologies derived from such knowledge – are traded in one form or
     another between innovative firms and other entities. Getting access to such
     knowledge has become a key part of the business plan of many such firms
     and although the boundaries of firms has shrunk as they access knowledge
     from elsewhere, the reach of their governance efforts have expanded, going
     increasingly beyond the firm boundary, to keep control of their new lifeblood:

                                  KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                              2. KNOWLEDGE FLOWS –    33
       Figure 2.1. Contributions to growth in global R&D, 1996-2001 and 2001-06
                                    In billion constant US PPP and %


      160                 12%
      140                 11%                             13%
      120                 10%                                                    Other non-OECD (2)
                           7%                             30%                    China
                                                                                 Other OECD (1)
        80                23%
                                                          13%                    Japan
                                                          13%                    EU-27
        40                                                                       United States
                          37%                             15%
                       1996-2001                       2001-2006

        1. Australia, Canada, Iceland, Korea, Mexico, New Zealand, Norway and Turkey.
        2. Argentina, Brazil, India, Israel, Russian Federation, Singapore, South Africa, Chinese
        Taipei. The statistical data for Israel are supplied by and under the responsibility of the
        relevant Israeli authorities. The use of such data by the OECD is without prejudice to
        the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank
        under the terms of international law.

           Thus, knowledge interdependence has driven actors to access and
       manage knowledge in new systems of “open innovation”, where ownership,
       or at least governance,1 of knowledge outside the traditional firm boundaries
       has become as essential as management of cash flow. There is a common
       misperception that closer governance of knowledge, which in biomedicine
       currently is principally through holding patents, is at odds with “openness”
       or collaboration. In fact, good governance of knowledge gives firms and
       other innovative entities the confidence to enter into collaborations, along
       the lines of the maxim “you can’t trade what you don’t own”. The terms of
       trade can be set according to strategy and to market and sector conditions. A
       good example of such “smart” governance is the growth in co-ownership of
       patents (Figure 2.2).
           Seen in this context, the development of KNMs are means by which
       knowledge can be strategically governed within open collaborative systems.
       That said, KNMs are poorly described and analysed and (at least in the life
       sciences) are a very recent phenomenon intended to deal with the relatively
       long-standing problem of translating knowledge into growth and other
       socially desirable outcomes.


         This chapter explores why knowledge flows are important to innovation.
     It defines what knowledge is, explains how it is generated and exchanged,
     and describes some of the inefficiencies in the biomedical innovation
     process. It also identifies some types of knowledge which have traditionally
     been held proprietarily but which are being targeted for broader access and
     use in biomedicine. How can we enhance knowledge circulation and use?
     How can firms engaged in biomedical innovation make the most out of
     knowledge interdependence? What are the main pressures within the bio-
     medicine sector for the creation of functional knowledge networks and
     markets (KNMs)?

        Figure 2.2. Growth in numbers of patents co-owned between businesses

     2 500

     2 000

     1 500

     1 000


         1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

The definition of knowledge2

         Knowledge is the theoretical or practical understanding of a subject.
     Knowledge can be “embodied” in a person, as his or her expertise and skills,
     or “embedded” in an organisation or a system (for example, in a corporation
     or a database like Wikipedia). Knowledge can be “tacit” (when someone is
     not aware of the knowledge s/he possesses or how it can be valuable to
     others) or “explicit”. Explicit knowledge can be “codified” so that it is more
     easily transmissible to many others. Such codified explicit knowledge
     (process descriptions, regulations and procedures, clinical study design,
     learning modules) is of particular interest for KNMs as it can more easily be
     shared broadly.

                                  KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                      2. KNOWLEDGE FLOWS –   35

           Some of the main features of knowledge are that its production is a step-
       by-step process: outputs from one step are inputs into the next (Aghion and
       Howitt, 1997). Knowledge production is multi-faceted and key outputs
       include things as varied as information, data, materials, know-how, methods,
       services and inventions (Collins, 1974). Knowledge is non-rival and non-
       excludable in use, such that one idea can be an input into many different
       follow-on experiments or research projects (Collins, 1974). There is
       uncertainty about the trajectory of knowledge use or its most productive
       path (Shane, 2001) such that the possible uses of knowledge once produced
       are hard to predict. For these reasons, a single organisation is unlikely to
       pursue all avenues of potentially valuable follow-on work.
           At least in the biomedical field, few innovations originate and are
       developed within one and the same firm. In modern industrialised economies
       the division of labour often means that a producer does not carry out all the
       steps necessary to transform raw materials into finished products inside the
       firm: it buys inputs and sells outputs to other users and producers along the
       supply chain. Contracts and IP rights facilitate these transactions. External
       sourcing is normal given that most firms do not have all the expertise or
       resources in-house to carry a product from bench to bedside. Markets
       facilitate the exchange of knowledge.
             Broadly speaking, there are three aspects of knowledge: i) its production;
       ii) its transfer; and iii) its application. KNMs can facilitate all three aspects,
       though perhaps the greatest attention of policy makers has been on the transfer
       of knowledge through knowledge transfer networks (KTNs)3 and other
       devices. Knowledge transfer is not new, of course, and includes practices as
       mundane as letters, discussions, apprenticeships, new hires, education,
       training, mentoring and conferences. Explicit codified knowledge is trans-
       ferred through articles, books, formulas, models, materials, databases, and
       patents. More recently, information and communication technologies (ICTs)
       have permitted the creation of knowledge bases, expert systems, knowledge
       repositories, and computer supported co-operative work. The diffusion of
       ICT (and in particular the Internet) makes possible a vast improvement in
       the speed and extent of information transfer, and this change is only
       accelerating with supercomputing, grid and cloud computing and the virtual
       testing of products. KNMs can increase the efficiency of knowledge dis-
       semination but also increase knowledge production by allowing broader
       access to intellectual assets and the ability to pursue multiple applications.
       Increasing access to intellectual assets allows further experimentation and


         Because of these changes, the organisation of research in the biomedical
     sciences is in flux. The biomedical sciences are now awash in information,
     with knowledge intensive resources widely dispersed and increasingly costly
     to maintain (Bolton et al., 2011). Over the past decades the producers and
     users of biomedical data have become far more diverse and have accommo-
     dated many new players globally. Biomedical research institutions are thus
     looking for new ways of working to store and exploit the vast amount of
     data creation and the associated increase in its potential users.

The new knowledge complex

         The mere production of knowledge does not guarantee its accessibility
     or use. One problem is secrecy, which limits available knowledge. Secrecy
     is used both by the commercial and academic sectors (Cohen and Walsh,
     2007). Even with disclosure, ensuring access to knowledge is not straight-
     forward. According to Joel Mokyr, “[t]he economic impact of new tech-
     nology, no matter how ingenious, can be realised only if the institutional
     environment is conducive and allows for the exploitation of inventions in an
     effective manner.” (Mokyr, 2007). In other words, the institutional context
     allowing access to knowledge matters (Murray, 2002; Murray and Stern,
     2008). And it is precisely this institutional context that has been changing
     recently in the life sciences.
         Fiona Murray argues that the set of formal and informal institutions that
     support the exchange of knowledge can be seen as a “knowledge complex”.
     Historically in the life sciences, the knowledge complex largely comprised
     informal peer-to-peer relations within a community with a shared culture
     and norms. But this “old world” knowledge complex is under strain on both
     the demand and supply sides. The scientific community is growing, with
     greater participation of industry in the development of relevant knowledge
     and new entrants, such as researchers from emerging economies (Figure 2.1).
     Supply-side strains are a product of the increasing availability, complexity,
     geographic and disciplinary fragmentation of knowledge as well as the cost of
     its exchange (e.g. in the United States, new requirements for material
     transfer agreements in the academic world in the wake of the Bayh-Dole
     legislation). Demand-side strains are a result of the increasing size and
     diversity of potential research consumers. Thus, knowledge interdependency
     comes with a significant increase in the complexity of matchmaking
     potential knowledge suppliers and those that might use such knowledge, if
     only they could efficiently source it.

                                  KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                     2. KNOWLEDGE FLOWS –   37

           According to Murray, the Old World system is breaking down and is
       being replaced with new institutions mediating knowledge exchange.
       Collective innovation is the norm in the biosciences, but the growing
       fragmentation of knowledge across institutions and disciplines has created
       an opportunity for new institutions to replace peer-to-peer exchanges
       (Powell and Giannella, 2010). These collective systems constitute the
       concept here of KNMs (Box 2.1).

         Box 2.1. Examples of collective systems (i.e. KNMs) for innovation in the
  Data registries and repositories
  e.g. NIH GenBank, The Cancer Genome Atlas, Global Biological Resources Centre
  Network (GBRCN)4. Virus Pathogen Bioinformatics Resource Centre,
  Platform technologies and tools
  e.g. Sage Bionetworks, BioBricks foundation, GSK medicines for malaria platform
  Consortia and public-private partnerships
  e.g. Innovative Medicines Initiative, Biomarkers Consortium
  Pools, clearinghouses and exchanges
  e.g. GSK Neglected Tropical Diseases pool
  Prizes, on-line auctions, brokers and citizen science
  e.g. Innocentive, Prize4Life, Foldit

           In short, new arrangements and institutions are emerging to support
       knowledge exchange in a more complex environment for R&D in the life
       sciences. That is not to say, however, that emergence of KNMs is either
       planned or co-ordinated. Granting agencies, to date, have preferred to invest
       in new research over the creation of institutions such as KNM. The institu-
       tions and arrangements discussed in this report are generally stand-alone
       projects, not designed within a policy systems approach. This may be a
       system failing.

Biopharmaceutical industry perspective on knowledge flows5

           Intellectual property protection is extremely important in the biomedical
       industry and is seen as being at the cornerstone of the R&D process. Patents,
       in particular, are recognised as critical to the translation of basic discoveries
       into novel diagnostics and therapies in the biopharmaceutical sector. Classic
       studies have shown that the pharmaceutical sector relies very heavily on


       patent protection in bringing innovations to market.6 A well-functioning IP
       system is a prerequisite to having an innovative and globally competitive
       biomedical industry. In addition to patents, the bio-pharmaceutical sector
       uses secrecy and proprietary information, such as compound libraries or
       clinical data, as an integral part of their business strategy. Both biotechnology
       and pharmaceutical firms actively license in and out technologies and form
       alliances to access external knowledge. In order to bring innovations to
       market, they license in technologies and collaborate with outside firms and
       organisations to access complementary knowledge as needed. Out-licensing,
       on the other hand, allows firms to earn a return without making additional
       investments in development. The broader a technology’s application (the
       more fundamental the innovation), the greater the incentives are to license it
       out. Given how dynamic R&D is in the health sector, open innovation
       strategies – by which is meant the flexible use of in and out licensing – are

     Figure 2.3. Percentage of new approved drugs based relying more than 50% on
                        externally derived technology, 1989-2004

   (N=10)   94%
                      82%        82%      80%
                     (N=22) (N=22) (N=10)

                                                   56%        55%
                                                  (N=18) (N=11)      50%     50%        47%
                                                                    (N=26) (N=18)
                                                                                                                    (N=12)     13%








































Source: Ceccagnoli et al. (2009).

                                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                       2. KNOWLEDGE FLOWS –     39

           A recent study demonstrates that new drug approvals at the major
       pharmaceutical firms rely extensively on externally derived technologies
       (Ceccagnoli et al., 2009). Figure 2.3 shows, for firms with more than 10 new
       drugs between 1989 and 2004, the percentage of new drugs in which more
       than 50% of patents attached to the new drug were not held by the com-
       mercialising firm. For nearly two-thirds of the companies, more than half
       their new drugs relied heavily on externally accessed technology.
           Arora et al. (2001) argues that the system of technology exchange is
       dynamic: markets for technology are large and growing. As the number of
       patents issued has increased rapidly since the early 1980s, so have receipts
       from licensing. The total United States market for technology – measured in
       terms of licensing revenues of patent rights – was USD 66 billion in 2002
       (Robbins, 2008). According to Arora et al. (2001), the distinctive feature of
       the last two decades has been the emergence of knowledge as a tradable
       asset (see Figure 2.4). It has spurred licensing and collaborations, and has
       created specialised technology suppliers, intermediaries and new technology

           Figure 2.4. Growth in receipts from international licensing of patents
                                             Billions of USD


  80                                                                            European Union

  60                                                                            United States





          Klein-Evans notes7 that from a biotechnology firm perspective, there is
     no problem with knowledge flows, accessing IP or creating freedom to
     operate. Market forces have driven resolution of even the most complex
     patent situations. The brake on bringing innovations to market is not access
     to technology but the costs and risks of development.
         Brian Kahin (University of Michigan, Computer & Communications
     Industry Association) adds that patents facilitate markets for technology.
     They are the answer to Arrow’s paradox that you don’t know the value of
     knowledge until you have it, but once you have it you no longer need to
     acquire it. With patents, you can have knowledge and evaluate it; but if you
     do not own or license the patent you cannot use it, or at least not the
     inventions identified in the language of the claims. Thus, patents facilitate
     the outsourcing of innovation by allowing patent owners to talk freely about
     their assets with potential suppliers, customers and collaborators.
          Viewed from an industry perspective, the problem is not so much the
     need to increase the flow of knowledge through more sharing. Rather, there
     is a sense that there is perhaps too much knowledge circulating too fast for
     any individual firm to effectively manage.8
         IP protection does not automatically provide a solution for firms or
     researchers overwhelmed by the proliferation of information sources. KNMs
     are formal mechanisms that increase the efficiency of sorting through
     potential sources of knowledge.

IP market failures and knowledge flows

         The patent system was conceived as a solution to secrecy and poor
     knowledge flow. Patents are important to firms because they provide trans-
     ferable exclusivity, and thus a mechanism for exchange. Even if there were
     no licensing, because patent applications are published knowledge would be
     disseminated. But firms also have strong incentives to publish in the
     scientific literature as a way of attracting collaborators and future employees,
     thereby further disseminating knowledge.
         Increasingly, however, there have been calls for more sharing of
     knowledge (Bolton et al., 2011). Some types of biomedical knowledge – for
     example, compound libraries, failed clinical trial data, toxicology data,
     adverse events, model organisms and disease state models – have typically
     been held proprietarily. Often, such knowledge is protected by trade secret
     rather than patents – which either are not available or are hard to enforce.
     Maintaining such secrecy can be a real challenge in the digital age. If made
     more accessible to multiple potential users, these resources could help
     prevent wasteful duplication of work, be put to new uses and possibly

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                    2. KNOWLEDGE FLOWS –   41

       combined with other information to further innovation. As Munos and Chin
       (2009) argue, “[s]haring could be the key that allows companies to access
       the vast creative, intellectual, and technological resources required to tackle
       the formidable challenge of turning the riches of the genome into a treasure
       trove of new treatments.” Opinion seems to be split over whether this
       underutilised knowledge should be available for free or whether a return on
       investment might be made.
           Much pharmaceutical innovation rests on science carried out in public
       research institutions although some of that research is funded by industry
       rather than the public purse. Nevertheless, the fact that the public science
       base contributes so much to biomedical innovation stokes the fires of those
       that advocate an open access norm. Others9 emphasise disclosure per se
       rather than free disclosure. From a firm perspective, the question is perhaps
       whether a market value is realisable for negative or otherwise underutilised
       data that is greater than any erosion of competitive position associated with
       disclosure. Knowledge markets that act as third party brokerages might help
       create opportunities to trade such knowledge, but so far these are in their
            The second issue facing the biopharmaceutical health sector is that
       despite strong intellectual property rights and the existence of vibrant markets
       for technology, there is clear evidence of stagnant productivity, measured as
       new drug approvals per year, and rising development costs.
            Ashish Arora and Brian Kahin both suggest the productivity problem
       could in part be due to a failure in the division of innovative labour, which
       might be the case if the market for technology (licensing, R&D contracts,
       etc.) is imperfect. Arguments that the IP system impedes knowledge circula-
       tion in the life sciences are not new. Arora noted longstanding concerns that
       broad foundational patents and patent thickets in the life sciences inhibit the
       circulation of knowledge and cumulative innovation. Foundational patents
       in biomedicine – for example, the WARF patents on stem cells or the
       Harvard Oncomouse patent – cover platform technologies with multiple
       applications.10 Restricted access to these types of patents would limit their
       applications in new fields and reduce follow-on innovation.11
           In addition, the spectre of patent thickets, ever present in genomics, is
       that they may prevent cumulative innovation. However, to date there is
       mixed evidence that thickets discourage academic research, or that firms
       have trouble negotiating freedom to operate. And the breadth of protection
       for bio-pharmaceutical patents has grown increasingly restrictive. However,
       the proposition still stands: if we were better able to identify and value
       knowledge assets, could market failures be avoided? Will new ways of


     identifying and accessing relevant knowledge assets allow society to extract
     more value from them?
          Meantime, biotechnology firms are an important source of innovation in
     the health sector (OECD, 2009); they are more nimble than the large
     pharmaceutical companies, but they rarely bring drugs from the lab bench to
     the bedside on their own, nor have they been profitable as a group over the
     last 20 years (Pisano, 2006). Can new approaches to sourcing knowledge
     improve the productivity of bio-pharmaceutical R&D?
          Jean-Paul Garnier, former CEO of GSK, argued in a famous Harvard
     Business Review article that the way to solve the productivity problem was
     “to morph big into small” (Garnier, 2008). R&D needs to be reorganised
     into small, highly focused groups headed by the leaders in any scientific
     field. Companies need to seek the best science, wherever it resides, inside or
     outside the company. The mistake pharmaceutical firms made was to
     assume that R&D was scalable, and could be industrialised and automated.
     Pharmaceutical research remains enormously complex and expensive, but
     the research base on which it is built should be nimble, varied and a mix of
     internal and external. Logically, then, as pharmaceutical firms increasingly
     rely on external sources of knowledge, organising that knowledge, searching
     for it, evaluating it, and being able to access it have become very important
     activities. And that is precisely where KNMs step in.


         Knowledge flow is fundamental to innovation in the life sciences. The
     current system of knowledge circulation is under pressure for a number of
     reasons. First, enormous amounts of data are being created every day by
     researchers in the public and private sector, which in itself complicates
     storage, analysis and use. Second, the number of potential users of these
     resources has grown; they also come from a broader range of institutions,
     cross-disciplinary boundaries, and are global, which makes a system of
     knowledge flow, based primarily on personal relations and trust, untenable.
     Third, the model of development for biomedicine is also changing. While
     bio-pharmaceutical companies scour external sources for new ideas and
     leads for the next generation of heath care products, they are increasingly
     open to participation in a variety of partnerships, consortia, and other
     approaches to generating and sourcing knowledge and innovation.
         Experts differ in their evaluation of whether there are significant barriers
     to the circulation of knowledge that impacts innovation. Some feel that the
     present IP regime has allowed the emergence of very robust markets for
     technology; others think that it has limits as a facilitator of knowledge flow
     and that its possible dysfunctions (e.g. thickets, blocking patents, trolls) can

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                    2. KNOWLEDGE FLOWS –   43

       impede the use of knowledge in some cases. There is division too on
       whether and how (under what terms) current underutilised knowledge could
       be made more widely publicly available and so useable by other innovators.
           Moreover, experts are divided as to the significance of new KNM. For
       some, these arrangements are neither new nor disruptive. For others, a
       significant new knowledge complex is in the making. We are already seeing
       a number of initiatives that seek to better organise and make accessible and
       interoperable the vast amounts of data, tools and know-how being generated
       in the biosciences. The expected benefits include a reduction in search and
       transaction costs, the creation of value from presently underused assets, and
       the broadening of the innovator base to new potential “problem solvers” and
       innovators. These new arrangements will influence the future pace and
       direction of R&D and might well change how the life sciences are funded
       and performed.
           The next chapters explore in greater detail what sorts of KNMs exist,
       and what their policy implications are.



1.       This concept of “governance” is not inconsistent with open access, so long as
         openness is a conscious and well planned strategic choice.
2.       This section draws on contributions made by Fiona Murray, Wolfgang Maass, and
         Greg Simon at the 2008 OECD Workshop on Knowledge Markets in the Life
3.       See for example,
4.       Global Biological Resource Centre Network,
5.       This section draws on discussions by Ashish Arora, Brian Kahin, Jonathan Klein-
         Evans at the 2008 OECD Workshop on Knowledge Markets in the Life Sciences
         as well as other sources.
6.       For example, Mansfield found that 65% of pharmaceutical inventions would not
         have been introduced if patent protection could not have been obtained: Mansfield
         (1986). See also work by Cohen et al. (1997). A more recent article is from
         Grabowski (2002).
7.       Jonathan Klein-Evans (MedImmune) comments at the Workshop on Knowledge
         Markets in the Life Sciences.
8.       Ibid.
9.       Bolton et al. (2011). See also the Royal Society website on science as a public
10.      For discussion of genetic inventions see OECD (2002). Stem Cell patent debates:
         see Murray (2007).
11.      Although it remains to be seen how the recent European Court of Justice ruling on
         Case C-34/10 Oliver Bruestle v Greenpeace e.V will impact on innovation in cell
         therapy and regenerative medicine.

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                    2. KNOWLEDGE FLOWS –   45


Aghion, P. and P. Howitt (1997), Endogenous Growth Theory, The MIT Press.
Arora, A., A. Fosfari and A. Gambardella (2001), Markets for Technology: the
      Economics of innovation and Corporate Strategy, MIT Press.
Bolton, G., M. Rawlins, P. Valance and M. Walport (2011), “Science as a Public
      Enterprise: The Case for Open Data”, The Lancet, Volume 377, Issue 9778,
      pp. 1633-1635.
Ceccagnoli, M., S. Graham, M. Higgins, and J. Lee (2009), “Productivity and the
      Role of Complementary Assets in Firms’ Demand for Technology
      Innovations,” Industrial and Corporate Change, Figure 1, p. 29, September.
      Available online at
Cohen, W.M., R.R. Nelson and J. Walsh (1997), “Appropriability Conditions and
     Why Firms Patent and Why They Do Not in the American Manufacturing
     Sector” Working Paper (Pittsburgh: Carnegie-Mellon University).
Cohen, W. and J. Walsh (2007), “Real Impediments to Academic Biomedical
     Research,” Innovation Policy and Economics, Vol 7, 2007.
Collins, H. M. (1974), “The TEA set: Tacit knowledge and scientific networks”.
      Science Studies 4:165-86.
Drucker, P F (1969), The Age of Discontinuity: Guidelines to Our Changing
     Society, London, Heinemann.
Garnier, J.-P. (2008), “Rebuilding the R&D Engine in Big Pharma,” Harvard
      Business Review, May 2008, Vol. 86 Issue 5, pp. 68-76.
Gordon, R. & Grant, D. (2000), Knowledge management of the management of
     knowledge: why people interested in knowledge management should read
     Foucault, in Clegg S Booth P Clarke T and Sominan F (eds) Deciphering
     Knowledge Management, New York, Springer-Verlag.
Grabowski, H. (2002), “Patents, Innovation and Access to New Pharmaceuticals,”
     Journal of International Economic Law 2002 5(4):849-60.
Mansfield, E. (1986), “Patents and Innovation: An Empirical Study” 32
     Management Science p. 175.
Marshall, A. (1972), Principles of Economics, 8th edition, London, Macmillan,
     first published 1890.


Mokyr, J. (2007),”The Market for Ideas and the Origins of Economic Growth in
     Eighteenth Century Europe,” [Heineken Lecture], in Tijdschrift voor
     Sociale en Economische Geschiedenis, Vol. 4, No. 1, 2007, p. 4.
Munos, B. and W. Chin, (2009), “A Call for Sharing: Adapting Pharmaceutical
     Research to New Realties,” Science Translational Medicine, Vol. 1,
     Issue 9, 2 December.
Murray, F. (2002), Innovation as Co-evolution of Scientific and Technological
     Networks: Exploring Tissue Engineering, Research Policy 3, 1389–1403.
Murray, F. (2007), “The Stem Cell Market – Patents and the Pursuit of Scientific
     Progress,” New England Journal of Medicine, Vol. 356 No. 23, 7 June,
     pp. 2341-43.
Murray, F. and S. Stern (2008). “Learning to Live with Patents: Assessing the
     Impact of Legal Institutional Change on the Life Science Community”.
     MIT Sloan Working Paper.
OECD (2002), “Genetic Inventions, Intellectual Property Rights and Licensing
    Practices: Evidence and Policy”, OECD, Paris.
OECD (2009), “The Bioeconomy to 2030: Designing a Policy Agenda”, OECD,
Pisano, G. (2006), Science Business: The Promise, the Reality, and the Future of
      Biotech, Cambridge: HBS Press.
Powell, W. W., and E. Giannella, (2010), “Collective Invention and Inventor
      Networks”, Chapter 19 in Bronwyn Hall and Nathan Rosenberg, Handbook
      of Economics of Invention, Elsevier 2010.
Quintas, P., P. Lefrere, and G. Jones (1997), “Knowledge Management:
      A Strategic Agenda”, Long Range Planning, 30 (3): 385-91.
Robbins, C. (2008), “Measuring Payments for the Supply and Use of Intellectual
     Property,” unpublished paper, 10 March, Table 9, p. 44, available at
Shane, S., (2001), “Technological Opportunities and New Firm Creation”,
      Management Science, 47(2), 205-20.
Stewart, T. (1997), Intellectual Capital: The New Wealth of Organisations,
      London, Nicholas Brealey.

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          3. ADVANTAGES OF KNOWLEDGE NETWORKS AND MARKETS –   47

                                             Chapter 3

            Advantages of knowledge networks and markets

      This chapter explains what specific economic, innovation, and health-related
      concerns are driving the widespread experimentation in the life sciences with
      novel knowledge networks and markets (KNMs) as well as how these
      motivate participants. Four major themes emerge. There is a common desire
      to: i) accelerate the health innovation cycle; ii) reduce the risk and costs of
      research and development; iii) translate scientific advances into products
      that better meet societal health needs; and iv) reduce health-care expenditures.


         The following sections provide a variety of perspectives (private, public
     and not-for-profit)1 about the incentives for participating in knowledge
     networks and markets (KNMs) and what policies different groups would
     like the government to enact in support of KNMs.

From open innovation to KNMs in drug development

         The challenges facing the pharmaceutical industry worldwide include
     the rising cost and length of drug development, the loss of patent protection
     on many of the major blockbuster drugs and the emergence of new paradigms
     in clinical care, especially the trend toward personalised medicine. These
     factors are forcing pharmaceutical companies to seek new business models in
     order to remain competitive. Pfizer, like other major pharmaceutical compa-
     nies,2 has actively cultivated an open innovation business strategy.
         Pharmaceutical companies are searching for a better, more efficient
     research model. Joe Fecko, Pfizer CMO, has noted that his company for one is
     entering into a wide variety of collaborations with other companies, academia,
     regulators and the government. These partnerships, consortia and entirely new
     types of organisation are all intended to increase the flow of knowledge into
     and out of the firm, manage that knowledge, and foster a leaner approach to
     innovation. Tapping into knowledge networks and markets gives the company
     access to a large pool of external ideas and innovations at lower costs and
     risks than traditional in-house product development. For example, Pfizer sees
     strong collaborations with academia as a way to drive personalized therapies
     that are emerging from the science base. Alliances are also designed so as to
     share the risks and costs of product development with partners. Ideally, this
     open innovation approach will increase the number of medicines Pfizer can
     move into clinical trials and ultimately into the clinic.
          Pfizer is currently involved in some 800 alliances across the full spectrum
     of research, development and commercial activities. In particular, the use of
     multi-player research consortia is on the rise. In 2009, Pfizer was involved in
     about 40 consortia and public-private partnerships worldwide focused on
     discovering and developing new medicines. Consortia provide a way of
     distributing across multiple firms the costs associated with pre-competitive
     research while reducing duplication of efforts. Examples of this sort of
     collaborative research are most common in oncology, but are expanding to
     other disease areas as well (see Table 3.1).

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          3. ADVANTAGES OF KNOWLEDGE NETWORKS AND MARKETS –                                                                                                                                                 49
                               Table 3.1. Consortia focus by disease area

                                                                                            Allergy and respiratory
                                                                    Genitourinary and sex

                                                                                                                                                           Infectious disease




 US biomarker consortium

 UK medical research council
 EU framework 7, innovative medicine

 Top Institute Pharma (Netherlands)

 BioWin (Belgium)

 Canceropoles (France)

 Medicatentos innovadores (Spain)

 Standardisation and harmonisation of
 imaging for clinical trials
 ADNI – Alzheimer’s disease
 neuroimaging initiative
 Ostoearthritis initiative
 Cardiac safety research consortium
 Imaging consortium for drug
 development (ICD)
 Mass insight collaborative imaging
 research center (CIRC)
Source: Feczko (2008).


          Consortia are also used to establish an early dialogue between government
     and industry about how new products and new approaches might be regulated
     and what methodologies and data will be required. In particular, new disease
     treatment paradigms (e.g. nanotechnology, gene therapies, regenerative
     medicine and cancer vaccines) require increased collaboration between parties
     to determine the best approaches and methodologies for clinical development.
     Consortia can build the knowledge base that allows a consensus to emerge
     amongst stakeholders on future clinical and regulatory pathways and assure
     patient safety. When government regulators are involved in such discussion
     at the outset of new fields (as in the Innovative Medicine’s Initiative, the
     Biomarker Consortium or Top Institute Pharma, as discussed below), their
     input helps reduce uncertainty about the development path lead innovators
         In addition to consortia and alliances, Pfizer is actively experimenting
     with different models of in-house research. In the past, R&D laboratories were
     bricks-and-mortar organisations staffed by employees. But increasingly the
     company is looking to other approaches (such as virtual R&D laboratories
     that bring together internal R&D capacity and external academic expertise)
     that work more flexibly, and to access the strengths of outside organisation.
     Pfizer does not do blue-sky research itself; its added value is translating the
     outcomes of research into practice. New R&D models could allow Pfizer to
     focus its comparative advantage.
          Despite much of the pharmaceutical industry’s embrace of open innova-
     tion, the concept of knowledge markets in life sciences is still in its infancy.
     Examples of new approaches to knowledge access and generation are more
     common in some areas, such as biomarker identification and validation,
     predictive pharmacology and toxicology, and novel clinical trial design. These
     sorts of initiatives tend to focus on advancing pre-competitive research,
     developing common approaches that will allow improvements in the safety of
     therapies (for example, improving toxicity predictions, creating models for
     testing toxicity) or agreeing on what constitutes proof of efficacy. Overall, the
     industry in general, and Pfizer in particular, is now sharing knowledge assets
     much more broadly than it was five years ago because access to KNMs
     through partnerships and consortia affords the opportunity for greater
     innovation, less cost and risk. Ultimately, this should lead to more medicines
     in the hands of patients in need.
         Looking forward, however, one can envision that increasing use of KNMs
     could have deep implications both for drug-development costs and health-care
     delivery. Drug development could benefit from stronger collaboration
     between academia and industry; novel and collaborative clinical trial designs;
     a dialogue with regulators to improve patient safety; and the safe sharing of
     data between health care providers, industry and regulatory authorities to

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          3. ADVANTAGES OF KNOWLEDGE NETWORKS AND MARKETS –   51

       speed the conduct of clinical trials and the regulatory process. At the moment,
       this sort of data sharing is rare.
           Ultimately, advances in the use of KNMs will improve clinical care.
       Pooling knowledge will help identify areas of medicine needing attention. It
       will also provide insights into fundamental aspects of diseases and
       conditions. KNMs could be a source of knowledge about how to leverage
       research findings to bring new medicines to patients more quickly. Ideally,
       KNMs would also channel insights from clinicians on what treatments and
       medicines their patients need.
            However, the value of KNMs will depend on how they are used, how
       many of them there are, and how they affect competition in the industry. It is
       possible to have too many consortia or other types of initiatives, which lack
       a clear purpose and do not deliver value. It is equally possible that a few
       (strong) consortia or other arrangements could standardise the industry and
       in fact reduce competition. Ideally, the growth in KNMs will redefine the
       pre-competitive research space, making it larger, and change the way we
       develop new medicines.

A network of pharmaceutical consortia

           KNMs are of interest to PPPs and other actors as well as the industry.
       For example, the Top Institute Pharma, which is a Dutch public private
       partnership with a mission to accelerate the development of socially
       valuable medicines is heavily involved in consortia building and collabora-
       tive research through various KNMs. TI Pharma was established to address
       the productivity problems plaguing the pharmaceutical industry and to
       reorient pharmaceutical priorities toward a public health-based research and
       development agenda. The institute brings together three partners - industry,
       academia and the government – that jointly fund and collaborate on projects
       that strengthen the scientific foundation of pharmaceutical research in the
       Netherlands. Specifically, its mission is to:3
            •    Create, through synergy, excellence in groundbreaking, cross-
                 disciplinary research, within the framework of Priority Medicines.
            •    Improve the efficiency of the entire drug development process, in
                 direct contact with and input from the regulator.
            •    Educate and train future generations of biomedical scientists.
           The TI Pharma research portfolio focuses on developing pre-competitive,
       “enabling” technologies, such as target validation tools, in silico modelling,
       animal models, and biomarkers. Projects exist in five therapeutic areas –
       autoimmune diseases, brain diseases, cancer, cardiovascular diseases, and


      infectious diseases (WHO, 2005). These projects are in areas deemed critical
      for pharmaceutical industry competitiveness, involve ground-breaking and
      cross-disciplinary research and have been chosen for the contribution they
      could make to improving the efficiency of the drug discovery and develop-
      ment process. Early input and dialogue with regulators is an important part
      of the TI Pharma approach, because it is key to reducing the time and cost of
      the approval process for novel medicines.
          TI Pharma has created over 47 consortia, involving 72 different partners.
      Consortia must have three partners, including at least one public and private
      entity. The public sector participants are Dutch universities, medical centres,
      and institutes like NKI (the Netherlands National Cancer Institute) and TNO
      (the Netherlands Organisation for Applied Scientific Research). The private
      sector includes multinational pharmaceutical companies and small- and
      medium-sized enterprises. The idea behind the consortia is to create small
      highly-focussed groups to work on a scientific and technical problem and to
      facilitate the ability of larger firms to source this external R&D.

                          Figure 3.1. TI Pharma structure and activities

               TI Pharma: Interaction & integration for inspiration & initiation

           Researchers:                                                            Project portfolio:
                                                42 project consortia
           - 200 PhD students                                                      - 42 projects
           - 150 Post-Docs                                                         - Partners & participants:
           - 100 Technicians                                                         20 academic
           - 50 Seniors                                                              35 industrial

                                                 Interaction & Integration

       Learning:                                  Inspiration & Initiation             Ventures:
       Education &                                                                     Development up to
       Training, EMRA (IMI)                                                            PoC at own risk

                  New programs:                                               ‘For fee’ / consultancy:
                                                   New activities
                  euSEND (and other                                           Broker to leverage expertise
                  future initiatives)                                         within consortium
                                        Debate: on Pharma relevant topics

Source: De Laat (2008).

                                              KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          3. ADVANTAGES OF KNOWLEDGE NETWORKS AND MARKETS –   53

           Top Institute Pharma acts as hub for these consortia (see Figure 3.1). TI
       Pharma works as an aggregator of information and best practices from the
       research projects and makes connections across the projects where
       necessary. In this sense, it seeks to channel the knowledge provided by its
       partners and stakeholders, functioning like the international headquarters of
       a large company with multiple different locations or like a high-tech cluster
       that accelerates the flow of scientific and business knowledge.
           The TI Pharma example illustrates one model for the transition toward
       networked drug development, in which multiple consortia (networks of
       researchers) exist simultaneously, with research being conducted in distant
       laboratories and the knowledge then pieced together to enable the develop-
       ment of new research tools or therapies. Knowledge, both background
       knowledge and any new knowledge created by the project, is shared amongst
       partners within a project, but not with the broader research community. In this
       sense, TI Pharma encourages open innovation through the proliferation of
       (closed) networks, but not open science or open access to research results.

       Making biomedical data accessible and useful through new
           FasterCures is a not-for-profit group whose mission is to accelerate the
       pace of discovery and clinical development of needed new therapies.
       FasterCures acts as a convener for non-traditional allies so that they can
       work together to create new business approaches vital to turning research
       breakthroughs into life-saving therapies. It also acts as a convener for the
       biomedical research community “to learn from one about new models of
       innovation and collaboration, to share best practices, exchange ideas, and
       find relevant tools and resources.”4
            Among the many challenges to accelerating the development of new
       treatments, FasterCures has argued for converting the vast stores of existing
       biomedical data into usable information and ultimately into transferable
       knowledge. The patient is an information resource: the “Rosetta Stone” of
       biomedical research. For example, a patient’s electronic health record
       provides data about treatment and outcomes that, if the data can be
       standardised and accessed, would be very useful information for research. In
       addition, patients can provide biological samples (e.g. tissue, blood) for
       research that might help unlock the connections between genes, proteins and
       the environment; but only if the patient allows such uses and the information
       collected on the sample is reliable. Finally, patient involvement in clinical
       trials makes it possible to evaluate whether new diagnostics, drugs, experi-
       mental medical devices and surgical techniques are safe and effective.
       Because patients are the key to improved information flow, FasterCures
       works to get patients more engaged in research.


          One of the biggest obstacles to creating a better flow of information, and
     more rapid creation of useful knowledge in biomedicine, is the way health
     data are produced and stored. Clinical health data are collected and stored at
     present in a way that strips them of context and reduces their utility. There is
     often no way to link to past personal and medical information, to add
     information from biospecimens, or to connect to clinical trial data. To
     transform data into knowledge will require a shift to patient-centred data
     tracking, and this will require new structures – both new infrastructures, new
     analytic tools and new governance systems. Patient-centred data systems
     should link: i) electronic health records, which describe a person’s past;
     ii) bio-specimens, which are collected to meet a current demand for
     diagnosis; and iii) clinical trial data and other analysis which ultimately help
     predict future impacts of treatments.
         Another barrier to using medical data is the lack of a common language
     across bio-medicine: it is hard to connect data on health history, treatments
     and health outcomes because they are collected in different formats and
     coded differently. The quality of much of the data is also questionable.
     Creating standards that facilitate interoperability is a concern as health
     information systems become more widespread.5 This issue will be addressed
     in Chapter 4.
         Better data is not enough, however. The research enterprise itself has to
     change from the current focus on understanding biology that, undoubtedly,
     has led to advances in our understanding of human health and disease but
     only produces new therapies as a by-product of basic research. The focus of
     research should be explicitly on finding cures for diseases. A number of
     reforms could shift emphasis to solving health problems, among which
     changing academic tenure and promotion policies, which lead to data
     hoarding and the reward of basic research discoveries over their applications.
         There are enormous barriers, cultural, technical and legal, to making
     biomedical data useful and accessible. For this reason, FasterCures helps
     build collaborations between patients, doctors and researchers that enable
     better data collection and knowledge sharing across the public and private
     sectors in order to accelerate research on particular diseases. FasterCures
     fosters the creation of networks. It does not itself, advocate for particular
     forms of collaboration or sharing, but offers a platform where parties can
     discuss how to make the ideals of open innovation a business reality. The
     KNMs developed here go well beyond the traditional boundaries of
     innovation – bringing in actors from across the entirety of the cycle, and
     building a social community around the innovation problem, importantly
     including the ultimate users, the patients themselves.

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          3. ADVANTAGES OF KNOWLEDGE NETWORKS AND MARKETS –   55

The promise of information technology and knowledge markets in the life

           As well as these examples of using KNMs for more targeted, challenge-
       based innovation, there is also interest in using KNMs as a means to help
       academia transition from a ‘craftsman’ model of knowledge exchange to one
       that fully makes use of the information and communications technology
       revolution. The US National Science Foundation (NSF), for example, has an
       important role in promoting inter-organisational relationships such as
       industrial-academic partnerships. The “classic” model is the standard dyad
       (e.g. one firm and one researcher or research group), which is the peer-to-
       peer craftsman model of knowledge transfer, but NSF, like many other
       similar bodies in other countries, also pursues a broader version of industry-
       university collaboration through research programmes, often involving more
       than one university, each with research projects in a given area, and a
       number of firms associated with each university research project. This
       model increases the opportunities for collaboration and exchange, but each
       relationship remains dyadic.
          Going beyond dyads is difficult because it requires negotiating more
       complex multi-party agreements. NSF thus has a real interest in KNMs
       because they provide new models for broader knowledge transfer across
       multiple organisations (academic, commercial, and not-for-profit) and new
       approaches to contracting between these diverse parties.
           KNMs can also help overcome disciplinary divides to knowledge
       exchange, which is another barrier the NSF, as well as other science funding
       agencies, is working to address. The NSF is looking to create access points
       across multiple scientific disciplines. Perhaps the largest such project is the
       National Nanotechnology Initiative, which has a USD 1.5 billion budget and
       spans multiple disciplines including biology, health and agriculture.
            Moreover, the NSF has strong interests in the preservation and con-
       servation of knowledge. They are working with academic scientists to find
       ways to use and re-use data and information, especially irreplaceable and
       costly data (e.g. the first email, video from first moonwalks, data from rare or
       unusual events like comets, hurricanes, earthquakes, and data whereby there is
       an ethical obligation to assure like experimental data using individuals). The
       NSF also recognises the importance of replicating data for validation or
       training; it seeks to reduce the cost of data replication by increasing ease of
       access to existing resources. This requires data-sharing protocols and means to
       preserve the data. Data preservation is also important to enable longitudinal
       studies (e.g. assessing the impact of the American Clean Water Act by
       reviewing water quality data over the last 50 years; understanding global
       climate change by accessing photos of glacier retreat over decades). The


     NSF recognises the importance of supporting international scientific partner-
     ships and collaborations on scientific data preservation, and is interested in
     expanding participation of scientists from poor or remote places in
     knowledge networks.
          In short, the NSF has as its goal to improve both the preservation of, and
     access to, scientific data. It is seeking ways to make sure the systems it
     funds are open, nimble, reliable and sustainable over decades. The NSF is
     looking for new approaches to data storage and use, ones that will not rely
     on government funding, will be technologically flexible and respond to user
     needs, and will advance the scientific enterprise for years to come. Others
     are thinking along similar lines – in the United Kingdom, for example, the
     Royal Society (the national science academy) is at the time of writing in the
     midst of an enquiry into science as a public enterprise6 – where many of the
     issues facing NSF are of relevance. KNMs could perhaps provide a test-bed
     for taking forward many of these efforts.


         KNMs are of enormous interest in industry, in government agencies and
     in the not-for-profit research organisations. While many OECD countries
     see a vibrant bio-medical industry as an important element of the knowledge
     economy, the current business model for the pharmaceutical industry is
     unsustainable if research costs continue to rise and productivity does not
     improve. OECD countries seek to put downward pressure on health care
     expenditures, including on new biomedicines. Nevertheless, there is demand
     for new or improved health-care treatment, especially for cost effective new
     treatments that meet identified health care needs. In short, the present
     system for the development of new medicines is feared bankrupt. There is,
     therefore, uncertainty over what sort of business models will deliver future
     health technologies. KNMs offer some hope of an alternative.
          The lure of forming collaborations and alliances and of finding outside
     sources of knowledge as a means to accelerate the R&D process and reduce
     its costs and risks is common across a very wide variety of actors. KNMs
     seem to offer several positive things: i) they increase data, information and
     knowledge flows across organisations; ii) they create a more efficient
     division of labour; and iii) they enable the production of new knowledge.
     For companies, KNMs offer the opportunity to distribute risk across multi-
     ple projects and organisations, reduce development costs, source the best
     ideas wherever those reside and have a more flexible architecture.7 For a
     government health agency, KNMs might be called for when there is a
     market failure and companies are not providing a needed new product or
     service. For a public research organisation, KNMs leverage research capa-

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          3. ADVANTAGES OF KNOWLEDGE NETWORKS AND MARKETS –   57

       bilities and accelerate the translation of basic research into clinically useful
       products (see Table 3.2).
           KNMs will not, however, alter the reality that improvements in health
       care proceed mostly through incremental innovation. And while finding new
       ways to source knowledge and new ways to collaborate are essential to
       innovating smarter and faster, bio-pharmaceutical companies remain essential
       to developing and delivering therapies. KNMs must still allow for adequate
       rewards to innovators; they do not substitute for them.
           More effective use of biological and medical data is also a big driver of
       the creation and use of knowledge networks and markets. For example,
       targeted therapies are premised on researchers having been able to access
       information about patients with similar (e.g. genetic, metabolic, lifestyle,
       health history) profiles. However, it is an enormous logistical and policy
       challenge to build a large-scale infrastructure to enable access to high-
       quality, contextualized, longitudinal data and information for health-care
       research and health-care decision-making. New collaborations, networks
       and consortia are seen as important channels for testing new approaches to
       the collection, integration and analysis of biomedical data.
            Making regulatory pathways more predictable seems to be a third
       incentive for participation in KNM. Collaborations and networks offer a
       trusted environment in which to have an open dialogue between innovators
       and regulators about how to evaluate new biomedical technologies. The
       dialogue is not only about what data and information will be needed to
       satisfy the regulatory process but also about how to collaboratively develop
       common standards and platform technologies. Both public and private
       sectors agree that this sort of discussion, and its outcomes, needs to be
       science-based and transparent. There is even preliminary discussion of how
       to further open up the regulatory process by making more information
       publicly available on the results of clinical trials.
           Research collaborations, networks for data and information access and
       exchange, dialogues about the regulation of next-generation medicines –
       these are all part of creating a stronger partnership among innovators and the
       government. It is possible, however, that the partnership has to be even
       broader and deeper. KNMs are seen as part of the solution to OECD
       countries’ health care crises, because they reduce costs, by pooling some
       resources and outsourcing key elements of R&D. But they also raise
       questions about the costs of innovation in health care, how health related
       R&D should be supported, and society’s willingness to support it. The larger
       question of how OECD countries resolve the tension in health innovation is
       being played out in some of the debates over what type of “open” science
       and innovation they want to support.

                               Table 3.2. Policy objectives for KNMs

 KNM policy goal          Specific focus on
 Create new knowledge through improved data and information use
                          Higher quality biomedical data and specimens
                          Interoperable biomedical and related databases
                          Improved transparency and broader access to databases
 Expand pre-competitive research base
                          Validated therapeutic targets
                          Animal, disease models
                          Predictive drug disposition and toxicology information
                          Develop agreed biomarkers and biosensors
 Promote innovation networks
                          Improve knowledge flows across organisations and firms
                          Refine division of innovative labour through new research models
                          Improve valuation methods for intellectual assets
 Lower the cost and risk of product development
                          Novel clinical trial methodologies and designs
                          Safe sharing of clinical trial data to speed regulatory process
                          Consensus on toxicology and safety data to be used
                          Biomarker validation and regulatory acceptance
 Improve health care quality
                          Advance pharmacogenetics and targeted therapies for patient safety
                          Make electronic health records accessible for research
                          Improve clinical trial design and participation
                          Increase transparency of health research outcomes

                                            KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          3. ADVANTAGES OF KNOWLEDGE NETWORKS AND MARKETS –   59

1.          This section draws on contributions from: Joe Feczko, Chief Medical Officer,
            Pfizer, United States; Willem de Laat, Managing Director, Top Institute Pharma,
            Netherlands; Greg Simon, President, FasterCures, United States; Sylvia Spengler,
            Program Director, Information Integration & Informatics Cluster, Computer &
            Information Science & Engineering Directorate, National Science Foundation
            (NSF), United States, and other inputs.
2.          See, for example, GSK and Pfizer op cit.
3.          From the TI Pharma website,
            mission.html, accessed on 1 December 2011.
4.          From the FasterCures website, at
  , accessed on
            1 December 2011.
5.          Data interoperability and exchange are problems are key challenges being addressed
            through the US Office of the National Coordinator for Health Information
            Technology as it implements the Health IT allocations of the 2009 American
            Recovery and Reinvestment Act.
6.          See
7.          Some of these ideas are proposed by Garnier (2008).


De Laat, W. (2008), “The TI Pharma Strategy”, OECD Expert Workshop on
      Knowledge Markets in Life Sciences, Washington, DC, 16-17 October.
Feczko, J. (2008), “Reduced Drug Development Costs and Improved Clinical
      Care”, OECD Expert Workshop on Knowledge Markets in Life Sciences,
      Washington, DC, 16-17 October.
Garnier, J.-P. (2008), “Rebuilding the R&D Engine in Big Pharma,” Harvard
      Business Review, Vol. 86, Issue 5, May, pp. 68-76.
World Health Organization (2005), Priority Medicines for Europe and the World,
      WHO, Geneva.

                                  KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                  4. THEORIES FOR BUILDING KNOWLEDGE NETWORKS AND MARKETS –   61

                                             Chapter 4

      Theories for building knowledge networks and markets

      This chapter discusses how advances in information technologies are
      informing the design of knowledge networks and markets (KNMs), drawing
      heavily on experiences and insights from the ICT sector before applying some
      of the concepts from that sector to KNMs in biomedicine. The chapter first
      describes how, theoretically, knowledge markets are structured and function,
      and how some market failures could be overcome through better mechanism
      design. The chapter then discusses the importance of interoperability of
      separate data resources in biomedicine, and how to make these resources more
      interoperable and useful to many researchers. Finally, the chapter proposes
      how to apply these theoretical approaches to KNMs design, and to create a
      framework for understanding real-world examples of KNMs in biomedicine.

 The discussion in this chapter draws on contributions made by: Wolfgang Maass, Director
 of Research, Centre for Intelligent Media, Furtwangen University, Germany; and Jonas
 Almeida, MD Anderson Cancer Center, United States.


          Knowledge networks and markets (KNMs) can be defined very broadly.
      Indeed, even a workshop can be construed as a knowledge market: an
      audience demands knowledge and speakers supply it. But in order to under-
      stand the novelty of the open research initiatives, it is important to have a
      more structured understanding of KNM.
          Knowledge markets have been well conceptualised in the ICT field
      where some of the questions posed have included: what is knowledge, when
      can it be traded through an electronic marketplace, how do knowledge
      markets work, what are the different types of knowledge markets and what
      sort of business models work for knowledge markets.
           Of particular interest are electronic knowledge markets, defined as “inter-
      organisational information systems that allows the participating buyers and
      sellers to exchange information about prices and product offerings” (Bakos,
      1991). Knowledge markets exchange the “knowledge” they hold or generate.
      To be electronically exchanged, that knowledge needs to be made explicit and
      converted to “digital information goods”, which is the digital representation
      of the information and its associated economic value.
           The process by which a potential user transacts for access to knowledge
      is illustrated in Figure 4.1. There are four stages to the process:
          1. The information phase: during which potential users search for and
             evaluate the utility of a possible knowledge object. Acquiring suf-
             ficient information about the knowledge object is relatively simple
             in retail markets but difficult in knowledge markets.
          2. The intention phase: when a potential user signals their interest to
             the owner or intermediary.
          3. The negotiation phase: when buyer and seller negotiate a contract
             that covers the price and terms of access.
          4. The execution phase: arranging for the transfer of product, which
             can occur either before or after payment, and possibly some service
             and support to the user.

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                  4. THEORIES FOR BUILDING KNOWLEDGE NETWORKS AND MARKETS –      63
                         Figure 4.1. Service level of knowledge markets

                                  Intention                  Negotiation         Execution

• Search                    • Signaling                  • Negotiation       • Contract
    • Catalogue                                              • Fixed price
                                                                             • Knowledge
    • Text mining                                            • Auctioning
                                                                               object delivery
• Evaluation                                                 • Exchange
                                                                             • Financial logistics
    • Web 2.0/                                           • Contracting
                                                                             • Legal advice/
                                                         • Validation          mediation
    • Rating
                                                         • Signing           • IPR clearing
    • Consulting
                                                         • Archiving         • Service and
    • Legal advice                                                             support
                                                         • Enforcement
    • IPR clearing
                                                         • Mediation

Source: Maass (2008).

            A digital information good (DIG) is a digital representation of an
       information object that has an associated economic value. DIGs are supported
       by coding systems that provide context for the data in the form of semantic
       self-descriptions, including content, functional elements, qualifying elements
       and economic elements like the legal terms and conditions for access and use
       of the information, its price, etc. (See Table 4.1 for explanations of what sort
       of information is included.) The point is that in addition to the “information”
       or knowledge to be traded, a certain amount of associated information about
       the good – its properties, the nature of the proposed exchange – is needed to
       specify its attributes in enough detail to allow potential buyers to make a
       selection (Malone, et al., 1987).
           ICT (and so all) knowledge markets display, in principle, a number of
       market anomalies that can impede the exchange of knowledge. First, there
       are information asymmetries in knowledge markets between the buyer and
       seller, when the buyer often has to make a purchase before she/he can fully
       evaluate the knowledge. DIGs, however, reduce information asymmetry
       because the semantic self-descriptions should signal the utility and content
       of the information being offered to a potential buyer. Also, in knowledge
       markets marginal costs tend to be near zero because it is inexpensive to
       create copies (otherwise known as “lossless copies”). Finally, there is no


       distinction between the copy and the original, what is called the “copy
       anomaly.” New ways to describe and to value information goods are
       necessary to help develop KNMs. This, as discussed later, is a particular
       challenge in biomedicine and the life sciences in general – valuation and,
       even more problematic, description of the value of life science assets in
       order that they may be efficiently traded, remains quite intractable. Some
       (for example, Brown, 2008) have begun to think about independent third-
       party brokerages for knowledge valuation. This theme is returned to in
       Chapter 6.
           For easier querying of and exchange, some computer scientists are using
       a prototype coding systems called digital knowledge object (DKO) that
       demonstrates how data and information in a knowledge market can be
       contextualised. A DKO has a knowledge content description as well as three
       types of associated meta-information: i) a description of the content type
       (e.g. multimedia specification, fields of application); ii) economic informa-
       tion and access terms, including pricing and legal terms and conditions of
       license; and iii) meta-information that provides more detailed information
       about the content (e.g. pre- and post-conditions of applications making use
       of content).

Table 4.1. Elements for the representation of knowledge goods in a digital knowledge object

Qualifying elements           –   Author(s)
                              –   Ontological relations (domain ontologies, protocols)
                              –   History of modifications
                              –   Fields of application
                              –   Rating, reviews and evaluations
                              –   Content type: by reference or by value
Functional elements           – Transitions
Economic elements             –   Types of applicable trading modes
                              –   Pricing
                              –   Legal terms and conditions (incl. trading and IPR)
                              –   Roles and rights
Content description           – Content
Source: Maass (2008).

                                       KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                  4. THEORIES FOR BUILDING KNOWLEDGE NETWORKS AND MARKETS –   65

            The primary advantage of semantically annotated DKOs is that they are
       easily exchanged, for example they can be exchanged over the web using
       any kind of semantic web application. This approach to knowledge
       exchange is lower cost and encourages the development of mash-up
       applications: applications that plug together small modular applications into
       useful configurations (e.g. Google). To maintain ease of access from a
       technological perspective, DKOs were developed such that they are
       consistent with open semantic representations1 and can be transferred using
       open protocols.2 DKOs are of interest because they make it possible to
       bundle knowledge together with meta-data about that knowledge, whereas in
       a more traditional “database” model knowledge is exchanged in large
       distributed systems and the context information has to be added (with
       difficulty) ex post.
           This approach to the digital exchange of knowledge is promising for the
       future of KNM, but there are a few questions about its feasibility. It is not
       yet clear whether DKOs, with their faceted interpretation of knowledge
       content, are broad enough to be used across a range of applications. The
       DKOs may need further standardisation, for example of the knowledge
       object representation format, of the corresponding vocabularies used, or of
       the exchange protocols and market services for knowledge objects in a field
       like biomedicine.
           KNMs will probably benefit from bottom-up approaches to design, as
       opposed to their top-down creation, in which designers à priori try to create
       a functional environment for knowledge exchange by populating databases
       or facilitating interoperability or anticipating the types of queries that will be
       asked. The success of bottom-up applications like Google teaches us that
       small but scalable approaches can be very powerful. We still do not fully
       understand how different communities will seek to use KNM, and indeed we
       are just beginning to see how they might be used by the biomedical sectors.
       Advances in information technology and computer science will help make
       KNMs more useful, versatile and tailored.

Typologies for knowledge networks and markets

           Nevertheless, there is already a diversity of KNMs that can be studied.
       Examples are found in e.g. engineering, consulting, medicine and art. The
       computer science literature distinguishes amongst types of KNMs along two
       main criteria: i) the nature of the trading of knowledge (commercial versus
       non-commercial); and ii) the nature of the community of user (open versus
       closed). Within an organisation, KNMs are usually non-commercial and
       closed; their purpose, for example, might be to improve product quality
       within the firm. KNMs can also be member-based, and either open or


      closed. Closed member-based KNMs include initiatives like patent pools
      and consortia that facilitate knowledge sharing amongst members. These
      exchanges are typically non-commercial, in the sense that the endeavour
      itself does not need to be profitable, but it could require members to pay fees
      and should add value to the commercial activities of its members. Open
      member-based KNMs support more general knowledge trading (supplied by
      members) to a broad range of potential users (e.g. scientists, patent brokers,
      the public), but this form of KNMs has yet to find a successful business
      model. Wikipedia is perhaps the best example of an open and non-
      commercial knowledge network or market that is not supported by public
      funds. Figure 4.2 below shows where different types of KNMs fall along the
      continuum of openness and the commercial nature of the transactions.

                       Figure 4.2. Knowledge-market dimensions

      Source: Maass (2008).

          KNMs also differ substantially in terms of how they work and their
      business models. In a study of five providers, Maass found that KNMs
      employ a range of pricing mechanisms (e.g. fixed price, negotiation,
      auction, reverse auction) and rely on different sources of revenue, including
      transaction fees, sales fees, subscription or membership fees, and advertising
      revenues as well as different payment mechanisms (Behrendt et al., 2005).
      Nor are the organisational structure and the role of participants – authors,

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                  4. THEORIES FOR BUILDING KNOWLEDGE NETWORKS AND MARKETS –   67

       reviewers, brokers, users, sellers, buyers, escrow agents – uniform across
       different KNM. Indeed, there may be multiple business models that emerge,
       for example open-source communities, consortia and consulting services.
           In any industry, KNMs are still at a very early stage in their develop-
       ment. The IT community is working on how to design the next generation
       electronic KNM. As yet, however, there is no agreed typology of the
       different “inter-organisational information system that allows the partici-
       pating buyers and sellers to exchange information about prices and product
       offerings,” or even more broadly of the variety of arrangements that expand
       the circulation of knowledge necessary for innovation to multiple potential

Data mining in biomedicine

           Turning to the life sciences, the magnitude and diversity of data
       generation in the life sciences over the past several decades is staggering.
       The challenge is organising this data, making it interoperable and
       searchable, in order to make it amenable to analysis (Deus, et al. 2008). Bio-
       informatics is the intersection of information and computer sciences and
       biomedicine. Since the late 1980s, starting with molecular biology, bio-
       informatics has enabled multiple groups to develop, manage and understand
       large-scale databases. But these data structures are fragmented. The web in
       theory permits access to multiple databases such as these and makes it
       possible to scale the dissemination of biomedical knowledge. Such broad
       access to data was impossible in the old model of scientific knowledge
       dissemination based primarily on peer-to-peer interactions.
           Computer scientists and bio-informaticians work on how to integrate
       data from multiple domains and disciplines, and how to manage property
       and access permissions when different groups own the data. Semantic web
       technologies bring the promise of meaningful interoperation between data
       and analysis resources. Algorithms, data structures and web computing can
       improve data mining when the diversity and magnitude of the data generated
       in biomedicine defies automated articulation among different efforts.
           Currently biomedical data is not easily made interoperable. There are
       many potential sources of data from many different fields (e.g. high
       through-put molecular data versus clinical data) and the technologies
       suppliers and users employ to develop information infrastructures and
       access tools are changing rapidly. It would be prohibitively expensive and
       time-consuming if all of these parties tried to make their data available
       through custom ICT solutions. Furthermore, the systems they would design
       to bring together the resources provided by these custom solutions would
       generate “untraceable” assets: data for which it is not possible to determine


      the full set of steps used in generation of the data, thus making the data
      themselves unreliable, which could corrupt research results. The pro-
      liferation of different custom data access solutions is both unmanageable as
      a system and not designed to evolve for future use.
          In order to ramp up R&D efficiency in the life sciences efforts need to
      be made to network ICT infrastructures and make them fully interoperable
      using biomedical knowledge engineering applications (Deus et al., 2008).
      Interoperability is defined as “the ability of two or more systems or
      components to exchange information and to use the information that has
      been exchanged” (IEEE, 1990) Interoperability has a syntactic and semantic
      component. Syntactic interoperability is the ability to get data, once told
      where the data are located.3 For example, the Internet has syntactic inter-
      operability because users can use URLs to access web content. Semantic
      interoperability is the ability to use data for a different purpose than the one
      that dictated its generation, which is harder to achieve.4 The ICT industry is
      developing enabling tools and standards that will make semantic inter-
      operability a reality in the next generation Semantic web.
          Why is interoperability important in bioinformatics? In data-mining,
      researchers are looking to discover predictive independent variables, which
      is akin to finding needles in a haystack. The data being analysed were
      usually created with another purpose in mind. Furthermore, critical co-
      variables are often found in other haystacks, other data sets. A lack of
      syntactic interoperability is simply inexcusable. But to do integrative
      research requires having knowledge of engineering environments where data
      and tools from multiple sources are accessible and fully interoperable. New
      technologies developed in bioinformatics can be used to weave a manage-
      ment model where multiple intertwined data structures can be hosted and
      managed by multiple authorities in a distributed management infrastructure.
           In bioinformatics, the demand for improved knowledge sharing amongst
      distributed researchers is coupled with the technologies that facilitate
      knowledge exchange. KNMs will become more common in bio-informatics,
      the only surprise is that they are not already more widespread. Semantic
      Web technologies have the potential to address the need for distributed and
      evolvable data representations that are critical for research in Systems
      Biology and translational biomedical research.

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                  4. THEORIES FOR BUILDING KNOWLEDGE NETWORKS AND MARKETS –   69


           In thinking about creating the IT infrastructures that support KNMs, the
       goal should be to create resources that allow knowledge to be used flexibly
       for unforeseen future R&D. One channel of investigation has been over-
       coming information asymmetries by providing more information to potential
       users about the content of data resources and the terms of access for data and
       information offered in a KNM. Another channel of research focuses on how
       semantic databases can be designed to facilitate interoperability and thus
       access to data and information in a network of databases for future unknown
            Will future investment in the development of biomedical databases be
       optimal? In general, people and organisations only invest in building
       infrastructures and institutions once they have a specific problem to be
       addressed. This might lead to the under-provision of biomedical databases
       that are a public good. And while academic organisations can usefully
       populate databases, they cannot be relied on to create the next generation IT
       infrastructure for biomedicine.5 They may need help at the inception with
       creating the proper infrastructure for semantic databases as well as support
       to maintain and, when the time comes, decommission databases6 at the end
       of their lifecycle. Indeed, it is important to consider both the creation and the
       demise of data systems. Tracking the user activity for a specific semantic
       database demonstrates there is usually a slow initial period of activity when
       the rules are being written, a sudden rise in the amount of data stored once a
       critical mass is reached and finally a slowing of further additions of data.
       Ultimately, the users no longer want the database to change but wanted it to
       remain as a reference. Users will rapidly mine the data for their needs; the
       challenge is creating a tractable, evolvable resource.
           Policy interventions should support interoperability so that tools and
       information are not tied to any specific choice of technology. The National
       Library of Medicine’s initiatives are good examples of how to increase
       access to data in a technology neutral way (see Chapter 5 for a further
       description). And it is critical too that public and private investment in
       database infrastructure be developed in a co-ordinated fashion to deliver
       optimal socio-economic benefit. Some of these points will be returned to in
       Chapters 6 and 7.


1.        For example RDF(S), RDFa, OWL.
2.        For example SOAP, HTML, SMTP.
3.        Syntactic interoperable requires, for example, the use of standards for data formats
          and communication protocols to exchange information.
4.        A limited example of semantic interoperability is provided by the way that dates
          are stored in systems consistent with XML (i.e. a mark-up language related to
          HTML). These systems can manage data information such that it can be
          recognised and used by other programmes for other purposes, provided that the
          dates are labelled with the appropriate XML tag and conform to standards for date
          formats. The ICT industry is working to extend this kind of functionality to other
          types of data by developing enabling data models, tools and software environ-
          ments. Efforts to make data sources accessible also rely in large part on the
          development and adoption of common “languages”, or ontologies, that map data
          structures to one another. These kinds of technological tools and standards enable
          the meaningful use of data across multiple programmes and systems.
5.        Even the incentives for scientists to populate database can be weak, leading to a
          collective action problem. See in particular: Nielsen (2012).
6.        The OECD Guidelines on Human Biobanks and Generic Research Databases
          (, for example, provide specific
          principles and best practices related to aspects of creation, maintenance and
          decommissioning of such databases.

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                  4. THEORIES FOR BUILDING KNOWLEDGE NETWORKS AND MARKETS –   71

Bakos, Y. (1991), “A Strategic Analysis of Electronic Marketplaces,”
      MIS Quarterly, Volume 15, No. 3, pp. 295-310.
Behrendt, W., A. Gangemi, W. Maass and R. Westenthaler (2005). “Towards an
      Ontology-Based Distributed Architecture for Paid Content,” Lecture Notes
      in Computer Science: The Semantic Web: Research and Applications,
      Volume 3532, pp. 163-175.
Brown, P. (2008), “Establishing a Clinical Trials Finance Facility: Finance and
     Licensing Proposal”, OECD Expert Workshop on Knowledge Markets in
     Life Sciences, Washington, DC, 16-17 October.
Deus, H.F., R. Stanislaus, D.F. Veiga, C. Behrens and I.I. Wistuba (2008),
      “A Semantic Web Management Model for Integrative Biomedical
      Informatics”, PLoS ONE 3(8): e2946. doi:10.1371/journal.pone.0002946.
IEEE (Institute of Electrical and Electronics Engineers) (1990), IEEE Standard
      Glossary of Software Engineering Terminology.
Maass, W. (2008), “A Knowledge Market Prototype: From Conception to
      Execution”, OECD Expert Workshop on Knowledge Markets in Life
      Sciences, Washington, DC, 16-17 October.
Malone, T.W., J. Yates & R.I. Benjamin (1987), “Electronic Markets and
     Electronic Hierarchies”, Communications of the ACM 30(6) pp. 484-497.
Nielsen, M. (2012), “Chapter 9: The Open Science Imperative,” Reinventing
      Discovery: The New Era of Networked Science. Princeton University Press:
      Princeton, pp. 187-207.

                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   73

                                             Chapter 5

           Case studies of knowledge networks and markets

      This chapter reviews several types of knowledge networks and markets (KNMs)
      that are currently in use in the life sciences and describes case study KNMs. For
      each case, the purpose, membership, the business model and type of knowledge
      exchanged in the KNMs in question is discussed.


          Knowledge networks and markets (KNMs) and the cases discussed here
      can be organised into the six broad (but not mutually exclusive) categories
      as introduced in Chapter 1:
          i) Data registries and repositories.
          ii) Platform technology and tool providers.
          iii) Research consortia, public private partnerships.
          iv) Pools, clearinghouses and exchanges.
          v) Prizes, on-line auctions, brokers.
          vi) Citizen science projects.

           The chapter concludes with a summary of some recurring challenges to
      the establishment of KNMs and to their ultimate effectiveness as catalysts
      for future innovation. It stops short of discussing policy options for their use
      for better biomedical knowledge management, sharing, and exploitation as
      these issues are discussed in Chapter 7.
          The cases presented here are intended to give the reader a better sense of
      the variety of arrangements that expand the circulation of knowledge to
      multiple potential users in order to accelerate innovation. The case studies
      are organised loosely by function.
          First, data registries, repositories and synthesisers are initiatives that
      focus primarily on aggregating useful data and information for research use
      by third parties. Examples include fully open data repositories (e.g. access to
      pre-publication datasets) and biobanks, or biological resource centres, which
      are open to qualified professionals. The Cancer Bioinformatics Grid and the
      NIH’s efforts to increase data access to clinical trial information are two
      very different examples of aggregating and making accessible biomedicine-
      related information resources.
          Second, platform technology and tool providers create added value to
      collected biomedical data and materials by developing tools, technologies
      and models that are of use to multiple third parties. The open BioBricks
      Foundation provides a registry of standard synthetic biology parts for free.
      Sage Bionetworks is creating new disease models, commercially but also
      through an open-access project. In 2010 GlaxoSmithKline announced that it
      would make the compounds in its library that inhibit the malaria parasite
      available for free, once again an important tool for neglected disease

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   75

           Third, research consortia and public-private partnerships (PPP) focus
       on pre-competitive research to reduce duplication of research within the
       group and/or to develop standards that will facilitate future research and
       development. PPP can be very large initiatives with significant public-sector
       funds, as in the Innovative Medicines Initiative, the SNPs Consortium, or
       smaller and focussed initiatives like Top Institute Pharma. There are also
       innovation networks and knowledge communities that are far smaller-scale,
       often with no public sector involvement, where the emphasis is on col-
       laborative approaches to learning and innovating, as in physician networks.
            Online auctions, exchanges, brokers include a wide variety of initiatives
       in which transactions are more arm’s length; there is no collaborative R&D
       element or common need to access particular data or tools that bind all
       participating parties together. Rather these institutions can be thought of as
       markets that facilitate the purchases of discrete goods, generating research
       efficiency by providing a simple contracting mechanism for research or a
       licensing mechanism for patented technologies. These initiatives include
       patent pools; IP clearing houses; licensing agents; IP auction houses; research
       brokers and innovation portals.
           Finally, citizen science projects are crowd-sourced, networked approaches
       to solving specific scientific or technical challenges. The participants are
       volunteers who agree to work on an identified challenge, and they contribute
       the fruits of their work to the group. Citizens have been recruited to work, for
       example, on classifying galaxies (Galaxy Zoo), bird population counting
       (Cornell Laboratory of Ornithology), protein folding (Foldit). Citizen
       science enlists the public in collecting large quantities of data or tackling
       technical problems. Volunteers are asked to follow guidelines and in some
       cases are trained. Usually, in citizen science projects the results are made
       publicly available.

Varieties of knowledge networks and markets

            There are multiple dimensions along which KNMs differ which are
       helpful in understanding the function and variety of KNMs as well as in
       articulating a typology. In particular the points on which initiatives may
       differ include:
            •    Nature of the community:
                        How open or closed the community is that can use the
                        arrangement (e.g. is it limited to consortia members or open to


                    Who are the participants in KNMs (e.g. firms, not-for-profits,
                    NGOs, civil society, individuals, government agencies, public
                    research organisations, universities).
                    What are the roles of different participants – e.g. authors,
                    reviewers, brokers, users, sellers, buyers, escrow agents?
                    Is individual participation voluntary, mandatory, or based on
                    the role an organisation plays?
          •   The nature of the knowledge exchanged:
                    What type of knowledge is produced?
                    Are contributions modular (specific problems or tasks) or
                    project based?
                    Where along the innovation cycles is the created knowledge
                    most useful?
                    Is the knowledge independently or co-operatively developed?
                    How open is access to intermediary or end products, i.e.
                    members only, made public?
                    What is disclosed and through what means?
          •   Business models
                    Is the KNMs primarily commercial, not-for-profit, or non-
                    What are its sources of revenue, its pricing and payment
                    Is the knowledge valued or priced?
                    What are the terms of access for knowledge validation,
                    accumulation or use?
          •   Transaction types
                    What are the sorts of transactions taking place within the
                    arrangement (e.g. knowledge sharing, licensing, mapping and
                    What is disclosed and through what means?
                    Are interactions between participants entirely virtual and
                    discrete or is there some form of co-operation or co-ordination
                    over time?

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   77

            •    Governance structures
                        Is the network entirely virtual, is it housed at or within an
                        institution, or is it a separate but legal entity?
                        How are the community’s activities defined from the bottom
                        up, or top down?
                        What is its governance structure?
                        What are the terms of access for participation in the community?
           The case studies of different approaches to creating KNMs below will
       be discussed in light of these differences in characteristics.

       Clinical trials: registration and results reporting as first steps toward
       data sharing1
           The US National Library of Medicine’s (NLM) mission is to increase
       access to medical information. The NLM is the world’s largest medical
       library and among its functions (which also include internal and external
       research support and training) the NLM provides biomedical information
       infrastructures for a variety of users and audiences, such as:
            •    Medline and PubMed Central – which gives access to abstracts and
                 citations of published biomedical literature.
            •    MedlinePlus and Genetics Home Reference – which provides
                 medical information for consumers.
            •    Genbank and dbGaP – a public database of gene sequences and
                 genome wide associations created for and by scientists.
            •    PubChem – a database of small molecules.
           The NIH also encourages data access and flow through its public access
       policy, which aims to improve access to journal articles resulting from NIH-
       funded research.
            All of the above are data registries. Of particular interest here, is the
       initiative at the NLM to expand, the world’s largest
       clinical trial registry. Clinical trials are research studies that test how well
       new medical approaches work in people. A study answers scientific
       questions about new approaches to screening for, diagnosing or treating a
       disease. Some clinical trials also compare a new treatment to an existing
       one.2 Clinical trials are mandatory for the regulatory approval of new drugs
       and certain classes of devices. The NLM was charged with giving public
       access to information about the purpose of, participants in, and location of
       clinical trials. The registry was established in 2000 and the United States


      government made registration of clinical trials mandatory for serious and
      life-threatening conditions.
           Over the course of the 2000s concerns about the integrity of the clinical
      trials system, fuelled in part by the Paxil and Vioxx scandals, led to further
      changes. In 2004 the International Committee of Medical Journal Editors
      (ICMJE) announced that only registered clinical trials could have their
      results published by ICMJE journals, thus providing a stick for compliance
      with registration (DeAngelis et al., 2004). In 2007 a new US Food and Drug
      Administration act increased the scope of mandatory reporting of clinical
      trials and introduced penalties for organisations that fail to comply.3
      Responsible parties are now required to report summary results of certain
      “applicable” clinical trials. Starting in 2008, NLM made available online the
      summaries of the results of each registered clinical trial.
           Clinical trials inform medical decision making, so easy access to a registry
      of trials and information about the research objectives and methods as well as
      the study outcomes is important both to health care providers and patients.
      Clinical trials require volunteer participants who put themselves at risk by
      agreeing to new, experimental treatments. Ethically, many believe that such
      risk taking is a public service and the quid pro quo of participation in clinical
      trials should be that knowledge derived from clinical trials should be
      accurately and transparently communicated by those undertaking the trial in
      order to advance medical care. Increased transparency in clinical trials may
      help recruit patients to new or on-going clinical trials. Transparency can also
      help institutional research boards to better evaluate the risks and benefits for
      patient volunteers of clinical trial research proposals. Finally, clinical trial
      reporting can support a variety of research and analysis initiatives.
          The expansion of required attention to interface
      design and quality control. The data need to be presented in a manner that
      can be interpreted correctly by physicians, medical researchers and patients.
      The NLM had a short timeline for development, and had to service a wide
      range of users, so it opted to stay with familiar models used in medical
      journal articles, providing information on the phase of the trial, the condition
      treated, interventions, recruitment details, pre-assignment details, reporting
      groups, participant flow and measures of primary outcomes.
           Clinical trials generate masses of important data about safety and
      efficacy. Very little of that information is ever published and even less data
      is made publicly available for use by other researchers. There is a continued
      push for “more systematic disclosure of clinical trial information,”4
      including better reporting of clinical trial protocols and outcomes as well as
      post marketing studies. Some have called for a system “to pool data as they
      emerge from various clinical trials of a medication and aggregate the

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   79

       information for a fuller picture of a drug’s harms and benefits” (Sanger,
       2009). The NLM is seeking to include adverse-event reporting information
       (i.e. information about reported side effects which may be linked to the
       intervention) in their clinical trials information service. The NLM also
       strives to create linkages among its data sources in ways that promote
       further scientific discovery. It plans to integrate the clinical trials database
       with other NLM databases and resources (e.g. PubMed, PubChem).
       Researchers would then be able to use a web interface to move directly from a
       clinical trial summary to an article on the subject of the clinical trial to
       molecular visualisation.
  is an online registry of federally and privately
       supported clinical trials conducted in the United States and around the
       world. The data is open to all interested users at no cost. The United States
       government mandates contributions to the registry, manages it and oversees
       the quality of the information. The knowledge is primarily intended for the
       use of clinicians and patients, since the nature of the information included is
       not suited as a research resource.

       caBIG Health: Making biomedical databases interoperable5
            Several major NIH/NCI initiatives can be thought of as precursors to the
       development of true KNMs. Through initiatives like caBIG and the BIG
       Health Consortium, the NIH and the NCI are working to advance personalised
       medicine and develop a research model that unifies basic discovery, clinical
       research and clinical care into a seamless continuum and results in improved
       clinical outcomes through the use of personalized medicine approaches.6 The
       hope is that by better using biological and clinical research results, NCI can
       accelerate the translation of discoveries into patient benefits.
            To promote personalised medicine, the NIH and the NCI are working to
       create an IT platform that brings together different communities in bio-
       medicine, allowing them to share and analyse data across different insti-
       tutions, different disciplines and at different points in the innovation cycle.
       The challenge is to develop a sufficiently scalable framework such that it can
       integrate these different categories of knowledge. Given the silo nature of
       biology and medicine, there typically is little knowledge sharing across areas
       of specialisation.
           The overall goal is to connect discovery, clinical research and clinical care
       in a synergistic way. The tremendous advances in paediatric cancer illustrate
       the power that can be generated by linking all participants in the health-care
       system. Acute lymphoblastic leukaemia survival rates have improved from
       5% in the 1960s to more than 85% today, and can be attributed to two factors.
       First, there were critical improvements in technology: cell karyotyping made it


      possible to identify different varieties of cancer, which permitted researchers
      to test the efficacy of different therapies in the context of different diseases.
      Second, childhood cancer is treated in a context which blends care delivery
      and clinical research. Almost every patient with a paediatric cancer is enrolled
      in a clinical research trial. As a result, researchers and practitioners are able to
      correlate experimental laboratory data with clinical data (treatment, history,
      pathology, outcome, etc.). Clinical data are used to continuously evaluate
      outcomes and this continuous evaluation allows researchers to develop and to
      refine evidence-based strategies at an individualised level. These strategies are
      implemented by care providers and improve quality by adherence to
      continuously evolving standards of care. NCI wants to expand this model of
      integrated clinical research and care.
          The National Cancer Institute’s caBIG initiative was developed to connect
      the cancer research community through a shareable, interoperable cancer
      knowledge infrastructure. caBIG was launched in 2004 and has been deployed
      in 56 NCI designated cancer centres and 16 NCI-designated community
      cancer centres; these centres are responsible for the treatment of 20 million
      patients. The initiative is also expanding internationally (e.g. in the United
      Kingdom, China, Latin America).
          caBIG is deploying standard rules and a common language to more easily
      share information. It needs to link a number of knowledge silos and make
      them interoperable, including data from basic research (including large
      volumes of genomics and proteomics data), regulatory approvals, and clinical
      care. Integration requires the development of a common “vocabulary” and an
      interoperable infrastructure for data that is meaningful across groups. caBIG is
      also working to build or adapt tools for collecting, analysing, integrating and
      disseminating information associated with cancer research and care
      (including security tools governing user access).
           The technological approach underlying caBIG is a services-oriented-
      architecture: different groups can do different things but are connected by a
      technological lingua franca. The network is distributed to allow sharing of
      knowledge and tools where necessary, and segmented so that user access is
      restricted to the tools and knowledge based on the role of the various users.
      User access guidelines are determined in accordance with caBIG’s frame-
      work for data-sharing. This framework was developed to ensure that data-
      access decisions are sensitive to the ethical, legal and social issues
      associated with research involving human subjects (e.g. patient privacy) and
      are used to develop the contractual terms associated with gaining access to
      certain types of data through the network.

                                      KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   81

           In short caBIG is much more than a data repository, it provides a
       platform to facilitate data access and sharing, in the hopes of creating
       federated databanks. It also develops tools, for example its Tissue and
       Pathology Tools, In Vivo Imaging Tools. In short, caBIG is a consortium
       with a large but closed community of clinicians and researchers from
       different cancer centres and other organisations who work collaboratively.
       NCI stewards caBIG. The fruits of the collaboration are made available to
       the research community as well as to other government agencies, and non-
       profit and commercial organisations. But it is hoped, also, that researchers
       will be able to use the caBIG platform and tools independently to advance
       cancer research and development.

       Access to genome wide association studies through a public private
       partnership: GAIN7
           The NIH, through its Foundation, creates public-private partnerships with
       other federal partners, industry, academia and the philanthropic community.
       The NIH Foundation is the sole entity authorised by the US Congress to raise
       private funds in support of NIH’s mission. The NIH Foundation has supported
       over 50 public private partnerships, including GAIN, and the Biomarkers
       Consortium discussed in the next section.
           The Genetic Association Information Network (GAIN) was a
       USD 26 million partnership between the NIH, the National Human Genome
       Research Institute, the NLM, Pfizer, Affymetrix, Perlegen Sciences, the
       Broad Institute and Abbott Laboratories. GAIN’s objective, completed in
       2007, was to make publicly available genome wide association studies for
       six common diseases (e.g. Attention Deficit Hyperactivity Disorder, diabetic
       nephropathy, major depression, psoriasis, schizophrenia and bipolar disorder),
       immediately and broadly. The genetic data for each disease were combined
       with clinical results to create a resource for genetic researches. The selected
       genotypes and phenotypes comprise an enormous data set which is made
       available to qualified researchers through the NLM’s dbGaP. By combining
       the expertise of the public and private sectors, the research process was
       significantly streamlined and results made broadly available.
           GAIN also developed an extensive set of data access, data-use and
       privacy policies. Key elements of these policies are summarised in Table 5.1,
       and they became the foundation for all Genome-Wide Association Studies
       (GWAS) across the NIH. The IT system that provides data access is set up
       to clearly communicate the data-use agreements multiple times in the
       process of data access. A data-use committee speedily reviews and responds
       to any infractions.

                   Table 5.1. GAIN data-access and data-access policies

 Policy type       • Policy
 Data quality      • Extensive analysis beyond peer review is conducted
 Privacy and       • Home institutions are responsible for compliance
 confidentiality   • Phenotypic data is double-coded
 Human subjects    • Oversight for initial participation resides at the local level
                   • Data use restrictions are enforced via the Data Use Certification and oversight by an
                     NIH-based Data Use Review Board
 Intellectual      • GAIN data is pre-competitive
 property          • Pre-computed associations and the Data Use Certification discourage pre-emptive
 Data access       • Partners are not given advanced access to data
                   • Controlled access to individual genotype and phenotype data is provided by an
                     online application process and the NIH-based Data Access Committee review
 Publication       • Contributing investigators get a nine month head start on submitting analyses of their
                     data for publication

       C-Path and the ILSI Biomarkers Committee: Toward the pooling
       toxicology data and drug screening tests8
           Two of the holy grails in pharmaceutical development are: i) the ability
       to determine the effectiveness of drug candidates early in the development
       process; and ii) the ability to predict any adverse effects of drug candidates
       before they cause harm to patients. Two consortia have tried to address these
       challenges by pooling data and research results.
           The ILSI/HESI Biomarkers Committee is a United States consortium set
       up in 2002 to advance the development and application of biomarkers of
       target organ toxicity (cardiac and renal).9 It included 15 pharmaceutical
       companies, several universities and both United States and European
       government agencies. The objective of the Biomarker Consortium is to
       develop “new, more sensitive indicators of cellular and tissue damage” for
       use in pre-clinical safety studies of medicines. Several such biomarkers have
       been identified for kidneys and two qualified as pre-clinical biomarkers by
       the FDA and EMEA in 2010.10 The Biomarkers Committee builds consensus
       across industrial, academic, and regulatory sectors about how to apply newly
       identified biomarkers of toxicity in risk assessments. It actively dialogues with
       international regulatory agencies in order to establish their data requirements
       for newly identified biomarkers to be used in regulatory reviews.

                                           KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   83

            While the group developing new biomarkers is a closed community,
       whose members must pay dues as part of the not for profit ILSI Health and
       Environmental Sciences Institute, the Biomarkers Committee requires that
       all of the tests developed be free of IP-encumbrances and that all research it
       conducts be published.
           A related consortium is the C-Path Institute Predictive Safety Testing
       Consortium (PSTC), established in 2006 to cross-qualify new and improved
       preclinical safety testing methods for drug-induced organ toxicities, which
       should improve the ability to predict a drug’s safety profile earlier in the
       development process. C-Path PSTC members pool pre-existing research and
       clinical data that had previously been held proprietarily by firms. “The
       results of precompetitive research are meant to be made publicly available,
       subjected to scientific scrutiny, and contribute to knowledge that improves
       the prospects for invention-based competition” (Woodcock, 2010). The
       consortium has developed more sensitive and specific markers for nephro-
       toxicity, which is helpful in assessing damage done by drug candidates in
       animal trials.
            It took the 17 companies involved about a year to create the C-Path
       Institute’s Predictive Safety Testing Consortium (PSTC) because of the
       difficulties in addressing IP concerns associated with the generation of new
       assays and biomarkers by the consortium. The consortium shares pre-
       existing samples and methods from rat, dog, and monkey studies to help
       validate biomarkers for use in animal studies. In principle, all of the data
       from partner firms is tested by all of the firms. In practice, companies focus
       on areas of expertise. The consortium is also working with the FDA and
       EMEA to develop a pathway for biomarker validation and approval.
           C-Path was created to pool data, share the costs associated with pre-
       competitive R&D, and increase the transparency and predictability of
       regulatory pathways so as to minimise the costs and risks of the regulatory-
       approval process. The initiatives demonstrate that there is an appetite for
       pre-competitive data sharing amongst certain groups, especially when the
       goal is to clarify the regulatory environment. These consortia can have
       different structures, and rules about participation and openness. But they
       also prompt questions about the limits or data pooling: what sort of
       knowledge is amenable to pooling and what companies are willing to
       concede as being “pre-competitive” R&D.

       The Biomarkers Consortium11
           The Biomarkers Consortium is another consortium associated with the
       NIH Foundation. Similar to C-Path, however, it’s goals are to: i) advance the
       discovery, development, qualification and regulatory acceptance of biomarkers;


      ii) conduct joint research in pre-competitive areas; iii) speed the develop-
      ment of medicines and therapies for detection, prevention, diagnosis and
      treatment of disease; and iv) make the consortium’s project results broadly
      available to the entire research community.
          The Biomarker Consortium’s contributing members include 24 for-
      profit firms and 31 not-for-profit organisations. Biomarker development and
      validation requires achieving a consensus, which consortia are well placed
      to build. There are thousands of candidate biomarkers in the literature, but
      practical use of the biomarkers requires validation, a challenging, time-
      consuming process that requires the development of consensus amongst the
      medical community. It is regarded as a pre-competitive activity, where the
      benefits of better biomarkers can be realised by multiple organisations
      seeking to address multiple classes of disease. The consortium also builds
      partnerships with organisations concerned with regulation, including the
      FDA, because regulatory pathways for biomarkers are still being refined.
          The Biomarkers Consortium organisational tree has three levels.
      Oversight is provided by the Executive Committee, which is comprised of
      representatives of the NIH and the Foundation for the NIH, the FDA, CMS,
      the life-sciences industry and the public. Four steering committees, in
      different disease areas, report to the Executive Committee.12 Each steering
      committee is charged with looking at the research landscape and selecting
      biomarkers for testing. Planning and execution of testing is done by smaller
      project teams who report to their committee.
          The Biomarker Consortium’s governance structure is designed to target
      high-impact areas of biomarker development and validation. Biomarkers are
      of interest to the consortium if they address a significant scientific need, and
      will result in significant improvement in the development, approval or
      delivery of care to patients. The consortium will undertake development
      only if the goals appear to be achievable in a specific time frame, with
      resources that have already been acquired or that are readily available. It
      seeks to maximise its impact by targeting opportunities that are not already
      being developed elsewhere, and that would uniquely benefit from the multi-
      stakeholder composition and approach of the Biomarkers Consortium. To
      ensure these criteria are met, the consortium employs a staged project-
      development process that calls for different types of oversight at different
           Since the Biomarkers Consortium focuses on pre-competitive research,
      it allows partners to contribute protected IP, which remains the property of
      the contributing entity. Project participants grant the consortium licenses for
      research use within each project. The Biomarkers Consortium requires parti-
      cipants to collaborate in a common research project. IP that is developed by

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   85

       the consortium is put in the public domain as much as possible to encourage
       downstream development. Similarly, raw data may be protected, or kept
       confidential; however, results are made broadly available as soon as possible
       after an appropriate review. Most general concerns are dealt with in the
       consortium agreement. Specific agreements define the exact terms for IP
       and data sharing for each of the projects undertaken by the consortium.
       Anti-trust violations require careful monitoring as the consortium develops,
            The development and management of PPP still is more of an art form
       than a science; however, there do appear to be some general requirements
       for an effective partnership. Chief among these, there must be agreement to
       a shared set of goals including a commitment to a clearly defined public
       benefit, transparent communication between partners, and a clear recognition
       of the distinct roles and limitations of each partner. Effective governance of a
       PPP requires the involvement of high-level leadership from all sectors, such
       that there is a balance between the public sector and the private sector. The
       collaboration requires an effective decision-making structure, which can make
       use of a trusted third-party intermediary when available. Partnerships are
       advised to develop policies that address the common IP and data-sharing
       issues as described in both the GAIN and Biomarkers Consortium. They are
       most likely to be effective when they have adequate financial and project-
       management resources, and a commitment to move quickly.

       The future of sharing access to chemical libraries13
            Chemical libraries are the foundation of drug-discovery programmes.
       These collections of chemicals are used to screen for molecules that act
       against a particular drug target. As seen in Chapter 2, the drug-discovery
       phase can account for over half of the total R&D costs. Ideally, the drug-
       discovery process investigates as many compounds as possible, including
       compounds with a diversity of structures.14 Access to a high-quality chemical
       library is vital. Large, high-quality compound libraries expedite the discovery
       process and lead to the identification of higher-quality candidate drugs and,
       thus, reduce the risk of clinical failures.
           There are two types of compound libraries. The first are the privately
       held compound libraries considered a key intellectual asset for pharma-
       ceutical firms and therefore not shared. These collections are made up of
       large sub-libraries based around individual templates; therefore, a library of
       two million compounds may only represent about 4 000 chemical templates.
       This is a concern for firms because if a template doesn’t bind to a target,
       none of the compounds associated with that template will bind to the target
       either. Increasing the size of these libraries by increasing the number of
       compounds is unlikely to be effective unless the number of templates is also


      increased. In other words, firms would benefit from access to more diverse
      compound libraries. To increase the diversity of drug-discovery resources,
      firms can turn to the second type of library: the commercially offered
      compound libraries. However, although many companies offer libraries of
      drug-like compounds, many of the compounds on offer are undesirable. The
      compounds tend to be too big and too lipophilic, and thus insoluble, to be
      viable as oral therapies. Analysis suggests that only about 10% of the
      compounds in these libraries are likely to be useful compounds for drug
      discovery. It is clear that the innovation capacity of the pharmaceutical
      industry as a whole is hindered by inadequate access to large, diverse
      compound libraries.
          In order to broaden the range of compound libraries used in drug
      development, one option would be to create shared compound libraries.
      These libraries could be created by consortia and would pool the best
      compounds from within the consortia members’ proprietary collections as
      well as from the collections available for sale. All consortia members would
      have access to the resulting pooled resource. The consortia could further
      seek to augment the shared libraries through large-scale chemical synthesis
      of new compounds using a design-driven approach.15 Health-care industry
      consortia might support the development of design-oriented chemical-
      synthesis libraries by pooling funds and putting out a tender for the design
      and the synthesis of the compounds.
           For shared compound libraries to become a reality, however, a number
      of issues need to be addressed:
           • Questions of IP ownership. The industry norm is to keep compounds
               proprietary, even when they are legacy data unlikely to provide any
               future value in terms of drug development efforts within the firm.
           • Legal issues associated with consortium development. These
               include the design of multi-party contracts in the context of anti-trust
          •   Preventing “free riders”. There is a concern that partners may
              provide consortia with the least-promising compounds in their
              libraries in exchange for access to the shared resource.
          •   Quality-control issues.
          •   Funding. Shared compound libraries will require funding and
              infrastructure to store, dispense and re-supply the chemicals. This
              infrastructure development may require private funding because
              public funders and foundations usually prefer to fund projects over

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   87

           The development of shared chemical libraries is most likely to occur
       amongst organisations that need to address difficult targets (e.g. protein-
       protein interactions) where privately held libraries are often inadequate.
       Consortia might be organised around gene families (e.g. kinases), or poly-
       pharmacology (e.g. kinases in cancer).
           To date shared compound libraries are rare in the pharmaceutical
       industry. The College of Life Sciences at the University of Dundee in
       Scotland has been a pioneer. Its Division of Signal Transduction Therapy
       (DSTT) is a collaboration between Dundee, the UK Medical Research
       Council and five pharmaceutical companies, who have joined forces to
       “accelerate the development of specific inhibitors of kinases and phosphatases
       for the treatment of disease [...]”16 DSTT aggregates compounds that
       previously were held as proprietary assets by pharmaceutical firms and
       develops new ones. According to the DSTT:
            “Participating companies share the right to utilise the DSTT's reagents,
            kinase profiling service, unpublished results and technical expertise.
            However, there is a fee-based structure to license the Unit’s IP, and for
            special services. Information or IP gained by using reagents, new
            technologies or information introduced to the DSTT by a company
            remains confidential to each company. […] Drafts of the Unit’s papers
            are placed on a closed website accessible only to each company.
            Company scientists visit Dundee three times a year for presentations of
            the latest results and discussions of mutual interest.”
           DSTT is a closed community, creating a common infrastructure that is
       useful in drug discovery. Contributors to the library have to invest in its
       creation, enter into contractual agreement with DSTT, and even pay for use
       of the knowledge created. The compounds contributed to the libraries are
       discrete and easy to specify, independently developed.
           Since there is a tremendous need to improve the productivity of health-
       care R&D and reduce its cost, it is possible these efforts to create such
       platforms will become more common, although their creation could happen
       slowly because they are formal collaborative structures that require a shift in
       industry norms and practices.

       OSDD: tools for drug development in India
           OSDD is an Indian consortium launched in 2008, by the Council of
       Scientific & Industrial Research (the premier R&D industry organisation in
       India), and funded in part by the Indian government. Its goal is to provide
       affordable healthcare to the developing world by creating a global platform
       where the best minds from different fields can collaborate to solve the
       complex problems associated with discovering novel therapies for neglected


      tropical diseases like malaria, tuberculosis and leshmaniasis. OSDD will
      aggregate biological and genetic information, and make it available to
      scientists to hasten the discovery of drugs using web 2.0 technologies. It
      already makes available a number of tools including: i) a Wiki-based
      genome annotation service; ii) the Computational Resources for Drug
      Discovery (CRDD), which is a platform of open source tools for drug
      discovery; iii) an integrative genomics map of Mycobacterium tuberculosis;
      iv) an Open Access Repository Document Repository; and v) a metadata
      archive and search engine for Open Access theses and dissertations. In
      addition, OSDD identifies “work packages”: well-defined scientific and
      technical problems that need to be addressed as part of larger drug-
      development projects managed by OSDD. The contributions researchers
      make to solving these work packages will be acknowledged and in some
      cases rewarded when a peer-review process agrees that a successful solution
      has been found. OSDD will award credits and appropriate prizes for the best
      solutions. The entire project is tracked and managed on line.

      Providing information about global health to a network of innovators17
          BIO Ventures in Global Health (BVGH) posits another approach to
      harnessing external knowledge in order to accelerate research in global
      health innovation. Since 2004, BVGH has focused on two barriers to the
      involvement of bio-pharmaceutical firms in global health challenges: “the
      financial realities that keep many innovator companies from focusing on
      neglected diseases with low market potential, and the lack of information
      about how they can apply their expertise and technologies to global
          BVGH’s first sought to identify what problems the bio-pharmaceutical
      industry could address which would have the biggest impact in the develop-
      ment of therapies for three neglected tropical diseases (i.e. malaria, tuber-
      culosis, trypanosomes). A 2008 BVGH study (BVGH, 2008) found that:
          •   R&D spending in global health is insufficient. Because of high
              attrition rates in drug development, there needs to be a much larger
              volume of drug candidates in the pipeline for the three neglected
              tropical diseases (NTDs) of interest if new drugs are to reach the
              market. Moreover, future development expenditures for NTDs will
              rise. The current pipeline for NTDs is mostly made up of re-
              purposed drugs. In the short term this saves time and money
              associated with regulatory approvals and manufacturing protocols,
              but in the longer term new therapies will require ramping up
              development expenditures.

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   89

            •    An innovation gap hinders NTD drug discovery. While foundations
                 and philanthropy funding has led to an unprecedented number of
                 late stage candidate medicines for NTDs, there is still insufficient
                 drug discovery funding for clinical development leading to a gap in
                 key technologies and drug discovery expertise and even portfolio
                 management capabilities.
            •    Discovering new NTD therapies is scientifically feasible. Bio-
                 pharmaceutical companies already have scientific tools for drug
                 discovery that can be applied to NTDs, such as knowledge about
                 targets and compound libraries. Where there are shared targets
                 between developed- and developing-world diseases, target platforms
                 can be applied to multiple diseases; which suggests that existing
                 bio-pharmaceuticals assets can be leveraged to help build the
                 pipeline and close the innovation gap between academia and
           With this framework for global health R&D, BVGH has promoted and
       developed financial incentives to engage biopharmaceutical companies in
       NTD R&D as well as information to compel and facilitate their involve-
       ment. Using this two pronged strategy BVGH hopes to leverage the existing
       bio-pharmaceutical knowledge base to expand the innovation pipeline for
       neglected tropical diseases.
           BVGH promotes the use of new financial incentives. The objective is to
       discover the set of circumstances that make it worthwhile for the bio-
       pharmaceutical industry to address the NTD market. In addition to helping
       to shape and promote the use of policies like the Priority Review Vouchers
       and Advanced Market Commitments, BVGH has also designed a new prize
       mechanism in which firms compete to develop urgently needed global
       health products.19 The Global Health Innovation Quotient Prize (IQ Prize)
       posits payments for reaching certain milestones of drug development. A
       pilot prize mechanism has been proposed for a diagnostic tool that would
       identify the cause of fever in children under five years of age. BVGH is thus
       making the market case for increased bio-pharmaceutical investments in
           The biggest barrier to industry involvement is a lack of information –
       about funding sources, about the identity of potential partners, and about the
       market opportunities associated with NTDs. BVGH seeks to address this
       situation by:
            •    Hosting meetings which support the development of new partner-
                 ships (e.g. Partnerships in Global Health Forum) and by facilitating
                 development of one-on-one partnerships.


          •   Drafting information tools, including: business cases that define
              markets, innovation maps to identify needs, new business models
              and strategies for development and delivery of treatments for NTDs.
          •   Launching the Global Health Primer, an online tool for researchers
              that showcases opportunities for research and product development.
              The Primer includes profiles of diseases, drug target and vaccines
              and technology profiles, the strengths or weaknesses, opportunities
              and risks for products currently in development.
          For example, Tuberculosis Vaccines: The Case for Investment is a case
      study that identifies a potential USD 1 billion market.20 BVGH shared these
      findings with vaccine companies, and four firms chose to further evaluate
      the tuberculosis vaccine opportunities. As a result, two tuberculosis vaccine
      programmes have been launched, and a TB biomarkers public-private
      partnership has been launched between a university and industry partner for
      the development of the relevant markers.
           BVGH is a broker, providing information, creating networks of innova-
      tors and even market maker. Its role is perhaps best described as making
      connections between underused resources – firms with expertise – and other
      partners in order to exploit new market opportunities. It provides information
      about innovation needs and gaps which is available to anyone interested.
      Challenges for these niche-brokers include finding an appropriate business
      model for their own organisations. BVGH is disinclined to move to for-
      profit status because that may jeopardise its status as an “honest broker”
      among other firms. One of the challenges to scaling operations like BVGH
      is in finding appropriate models for developing external partnerships. Right
      now BVGH helps other firms develop one-to-one partnerships, and it is very
      time-intensive work. Scaling the operation might require finding another
      model for both generating the information resources and developing partner-

      Collaborative knowledge networks for physicians21
          Some physicians are “independent inventors,” they develop new treat-
      ments, procedures or devices. There are a number of collaborative knowledge
      networks in medicine that are designed for or by physicians in order to learn
      from communities of practice and create an evidence for the identification of
      best practices.
          Independent inventors are an important source of breakthroughs in many
      industries. They are less likely to be entrenched in a given field and thus are
      not constrained by common assumptions within the field (i.e. find it easier to
      “think outside the box”). They may also benefit from having the autonomy to
      explore solutions unavailable to those more fettered by corporate strategies or

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   91

       problem-solving paradigms. Independent inventors create a higher level of
       knowledge diversity, which is associated with the discovery of greater
       numbers of high-value inventions.
            To explore the impact of independent inventors in the medical equip-
       ment industry, Lettl et al. (2009) looked at the knowledge base on which the
       patents of independent and corporate inventors was based. There were seven
       times more corporate than individual medical equipment patents in the
       sample. Technological impact was defined as the number of times a patent is
       cited in later inventions (i.e. patent citation measure). Corporate inventions
       were cited more frequently, which is evidence that they had higher
       technological impact. However, when one accounts for the characteristics of
       prior technical knowledge, independent inventions can have as large an
       impact as corporate ones. However, individual inventions (90% of whom
       were physicians) had technological impacts equal or higher to corporate
       counterparts only when they exhibited a high degree of specialisation
       (i.e. the patent cited many references in the same technological area, where
       knowledge is cumulative) and a low degree of diversity (i.e. the patent cited
       few references from other technological areas, probably because independent
       inventors are less able to absorb prior knowledge from other fields). In other
       words, independent inventors can even outperform corporate inventors when
       they make extensive use of existing knowledge in a particular technological
           How can independent inventors leverage technological specialisation?
       Knowledge networks and markets help physicians as independent inventors.
       Networks are important for independent inventors because through them
       they acquire state of the art knowledge and learn about trends; they get
       feedback on their ideas; they collaborate with others and help filter and
       select the most promising ideas; they gain legitimacy in communities of peer
       specialists; and their ideas and solutions are more quickly disseminated. The
       peer-to-peer knowledge exchange in networks accelerates innovation and
       improves professional practice. This creates a virtuous cycle in which the
       KNMs supports the development of individuals, as both practitioners and as
       potential inventors, and this growth in expertise is fed back into the
       community through the network. In sum, community embeddedness enables
       independent inventors to gain access to processes and resources which
       corporate inventors typically find within their organisation.
           One example of a learning network for surgeons is the AO Foundation,
       a Swiss-based not-for-profit founded in 1958 by surgeons who wanted to
       evaluate scientifically the use of a new surgical technique (internal-fracture
       fixation) in the hopes of refining it and improving patient outcomes.22
       Today, the AO Foundation is the world's leading knowledge organisation in
       osteosynthesis. It brings together surgical practitioners, research and


      industry for the benefit of patients. Its membership includes more than 5 000
      surgeons in more than 100 countries.
          The Foundation supports different activities for surgeons committed to
      the study and practice of the field of trauma and musculoskeletal surgery. It
      supports a clinical investigation and documentation by sponsoring indepen-
      dently conducted clinical studies that yield evidence based knowledge; it
      supports exploratory research through a network of scientific professionals
      that focus on truly novel approaches and theories in two fields of musculo-
      skeletal science; it provides start up grants for basic, pre-clincal and clinical
          In 2006, the was launched. It provides the participating
      surgeons with access to an online library with scientific papers, videos and
      descriptions of procedures. It is different from a database in that the
      knowledge assets are offered in an “evidence-based” format that guides
      surgeons through similar cases and provides multimedia supports. AO
      Surgery provides the complete surgical management process from diagnosis
      through to aftercare for all fractures of a given anatomical region. The site
      may evolve to include real-time support for surgeries. Currently, it has
      20 000 visitors a month. Contributors to the knowledge resources on the site
      include more than 50 of the world’s most renowned surgeons from at least
      20 countries.
           A second physician network was launched by Syndicom, a company
      whose mission is to enable communities and companies to seamlessly
      collaborate and share knowledge.23 Syndicom’s flagship product was
      SpineConnect, a platform that enables spine surgeons to collaborate on
      difficult and unusual cases.24 Surgeons in this online community post cases,
      research and problems, looking for insights and technologies from their
      colleagues. Industry representatives can also post exemplary cases about
      their technologies, and use data on the communities of practice to devise
      novel approaches to treatment or to devise technological solutions that
      address gaps in the current product market. To date it has collected over
      1 600 cases and 6000 reviews by over 1 000 members from over 30 different
      countries. Syndicom has since launched two other communities of practice –
      ArthroplastyConnect and TraumaConnect.
          These two types of networks, the not-for-profit AO Foundation and
      Syndicom’s for profit communities of practice for surgeons demonstrate
      there is a demand for networks to facilitate knowledge exchange in the
      medical community, both to address the learning needs of practitioners but
      also to promote faster innovation and uptake of best practices.

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   93

Online auctions, exchanges and brokers

       Tapping into a global network of innovators25
           InnoCentive is a knowledge market that links solution seekers to a
       global community of “solvers” through the online posting of challenges.
       Challenges are discrete (often, but not exclusively, technical or scientific)
       problems that customers generate, whose solution is associated with a
       monetary prize. The challenge once formulated is posted by Innocentive to
       either a select group (e.g. to a company’s internal staff, to clients) or to the
       global community through its website.
           InnoCentive was launched in 2005 as a spin-off of Eli Lilly.26 Some of
       the first group of solution seekers were large corporations, but solution
       seekers now also include foundations and governments. More than 1 300
       challenges have been posted to its Global Solver Community, in addition to
       hundreds of internal challenges. InnoCentive’s Solver community now
       consists of more than a quarter of a million individuals from nearly 200
       countries (about 40% of the solvers are from China and 23% from the
       United States). Between 20% and 30% of the solvers have PhDs.
           InnoCentive is a commercial “spot” market, which strives to maintain
       low transaction costs. The seekers and solvers do not negotiate. Instead,
       there are defined terms of engagement and online solver agreements. Seekers
       pay for solutions using a “bounty hunting” business model: the solver posts
       the problem and the bounty amount that they will pay for an adequate
       solution. This model distributes risk across the marketplace. It works because
       of self-selection of problems by individuals. Individuals post solutions, and
       typically opt to solve problems that they know they can solve easily or that
       interest them. By accessing such a broad solver community, this market model
       can tap into enormous diversity and does so better than other open innovation
       approaches, such as outsourced R&D or collaborations.
            InnoCentive provides the web-based infrastructure linking solvers, and
       its employees assist the seekers in each step of the process:
            1. Dissecting big problems into small ones so that the solver community
               can address them.
            2. Finding solutions, where submissions have included both documents
               and materials, which demonstrate that the solver community is willing
               and able to do laboratory work.
            3. Separate adequate solutions from inadequate solutions.
            4. Integrate solutions.


          Following the InnoCentive example, one could envision a pharma-
      ceutical industry of the future, which is comprised of the following three
      intersecting markets:
          •   A finance market that provides money in exchange for equity.
          •   An invention market that provides ideas in exchange for equity/
          •   A reduction-to-practice market that provides work in exchange for
           The first market represents business as usual. InnoCentive has
      demonstrated the feasibility of the latter two markets. All that remains is to
      link the three together with partners that provide other necessary skills and
      personnel. This “ecological” or “hive” model of pharmaceutical develop-
      ment could significantly reduce the costs of development because it
      distributes the costs of failed solutions across the solver community. In the
      bounty-hunter model, solution seekers pay only for solutions. Solution
      solvers accept the risk that they will generate a failed solution (and receive
      no compensation) because they self-select to solve problems and thus are
      able to choose to attempt only problems that interest them or that they have
      to solve for other purposes. By permitting lower cost development, this
      model of the future should also be able to address orphan drug development
      and neglected diseases, especially since it provides incentives for addressing
      problems that probably would not result in academic publications
      (e.g. methods, manufacturing). Worldwide healthcare needs more than good
      medicines, but this could be a good start.

          The KNMs described above illustrate the growing significance of new
      models of precompetitive research pooling, shared research infrastructures,
      knowledge sharing, and valued knowledge exchange in the life sciences.
      These initiatives are part of the vanguard of KNMs and can help policy
      makers understand the different types of institutions that are being created to
      better access and manage knowledge so that it can be more rapidly used to
      advance research or to develop much-needed new medicines. The case
      studies exhibit important differences in their purpose, structure, membership
      and business model. However, they are all devoting considerable resources
      to the design of governance arrangements, data-sharing policies, project-
      management protocols and other organisational design elements.

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   95

            At this point concerns about how to structure KNMs focus on:
            •    The management of IP rights and associated policies, especially in
                 collaboration with universities (e.g. university conflict of interest
            •    Financing, logistics and operations – for developing, maintaining
                 and sharing knowledge.
            •    Governance structures that allow the KNMs to make appropriate
                 decisions and provide evaluation of continued utility.
            •    Addressing liability issues, and antitrust regulations.
            It may be possible to reduce the overhead required for future efforts by
       documenting and tracking the success of these efforts so that such future
       initiatives can use the best approaches as templates.



1.        This section draws in part on a presentation at the 2008 OECD Workshop on
          Knowledge Markets in the Life Sciences by Jerry Sheehan, Assistant Director for
          Policy Development National Library of Medicine, United States.
2.        See MedlinePlus website:,
          accessed on 10 December 2011.
3.        US Public Law 110-85 (
          Title VIII, Section 801.
4.        See comments from Rob Logan, Ph.D. senior staff National Library of Medicine,
          at, accessed on
          10 December 2011.
5.        This case is based in part on comments made by Ken Buetow, Chief Information
          Officer at the National Cancer Institute, Director of the Center for Bioinformatics,
          United States at the OECD 2008 Workshop on Knowledge Markets in the Life
          Sciences. It also draws from the caBIG website at:,
          accessed on 5 January 2012.
6.        See Chapter 3.
7.        This section draws in part on a presentation at the 2008 OECD Workshop on
          Knowledge Markets in the Life Sciences by David Wholley, Director, the
          Biomarkers Consortium, the Foundation for the National Institutes of Health,
          United States.
8.        This section draws in part on a presentation at the 2008 OECD Workshop on
          Knowledge Markets in the Life Sciences by Federico Goodsaid, Associate
          Director, Office of Clinical Pharmacology, US Food and Drug Administration
          (FDA), United States.
9.        ILSI Health and Environmental Sciences Institute (HESI) is a non-profit scientific
          organisation that brings together scientists from around the world from industry,
          regulatory agencies and other governmental institutions, academia, and other
          research organisations. See the ILSI/HESA website,
, accessed December 2011.
10.       See the ILSI website,
, accessed December 2011.
11.       This section draws in part on a presentation at the 2008 OECD Workshop on
          Knowledge Markets in the Life Sciences by David Wholley, Director, the Bio-
          markers Consortium, the Foundation for the National Institutes of Health, United

                                     KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                          5. CASE STUDIES OF KNOWLEDGE NETWORKS AND MARKETS –   97

12.         The steering committees focus on cancer, inflammation and immunity, metabolic
            disorders, and neuroscience, respectively.
13.         This section is based in part on a presentation at the 2008 OECD Workshop on
            Knowledge Markets in the Life Sciences by Paul Wyatt, Director of Drug
            Discovery, University of Dundee, United Kingdom.
14.         Productive screening collections need not be as large as combinatorial models of
            chemical compounds suggest (i.e. these models can suggest potential compounds
            number from the hundreds of thousands, to the millions, or even more). While the
            number of compounds is huge, drug development focuses on those compounds
            that are most likely to bind to biological targets and be orally bio-available, that
            are consistent with good pharmacokinetics, and are unlikely to generate toxic
            responses. As a result, the number of compounds of significant interest is not very
            large relative to the range of potential solutions.
15.         A design-driven approach to chemical synthesis uses chemo-informatics to identify
            drug-like templates and side chains using open-source databases such as Starlite.
16.         In particular, these chemicals are important for rheumatoid arthritis, diabetes and
            cancer. See the Medical Research Council at
   For more information about DSTT see
            the websites of the College of Life Sciences at the University of Dundee, at
  ; and Bio-Dundee at
   Accessed on 10 December 2011.
17.         This section is based in part on a presentation at the 2008 OECD Workshop on
            Knowledge Markets in the Life Sciences by Wendy Taylor, Founder and Vice-
            President of Strategy and Operations, BIO Ventures for Global Health, United
18.         See BVGH website:, accessed in January 2012.
19.         For more information on both the PRV and AMC incentives see:
   See also OECD (2009).
20.         For access to the report, see
21.         This section is based in part on a presentation at the 2008 OECD Workshop on
            Knowledge Markets in the Life Sciences by Iwan von Wartburg, Managing
            Partner, iploit AG, CH; Senior Research Fellow, University of Hamburg, Germany.
22.         For more information about the AO Foundation see its website:
  , accessed in
            January 2012.
23.         Syndicom was founded in 2000 by three American academics: Raymond Miles
            (University of California, Berkeley), Charles Snow (Pennsylvania State University)
            and Grant Miles (University of North Texas).


24.       See the Syndicom Spine-Connect website:
, accessed in January 2012.
25.       This section is based in part on a presentation at the 2008 OECD Workshop on
          Knowledge Markets in the Life Sciences by Alph Bingham, Co-founder and
          Member Board of Directors, InnoCentive, United States.
26.       See the InnoCentive web page for more details on its functioning:
, accessed in January 2012.

BVGH (Bio Ventures for Good Health) (2008), Closing the Global Health
    Innovation Gap: A Role for the Biotechnology Industry in Drug Discovery
    for Neglected Diseases, BVGH, Washington, DC. Available at:
DeAngelis, C., J.M. Drazen, F.A. Frizelle, C. Haug, J. Hoey and R. Horton (2004),
     Clinical Trial Registration: A Statement from the International Committee
     of Medical Journal Editors. Ann Intern Med 2004;141:477-8.
Lettl, C., K. Rost, I. Von Wartburg, (2009), “Why Are Some Independent
       Inventors 'Heroes' and Others 'Hobbyists'? The Moderating Role of
       Technological Diversity and Specialization.” Research Policy, 38,
       pp. 243-254.
OECD (2009), Coherence for Health: Innovation for New Medicines for
    Infectious Diseases, OECD, Paris.
Sanger, N. (2009), “Public Database is Urged to Monitor Drug Safety,”
      The New York Times, 24 November,,
      accessed on 6 January 2012.
Woodcock, J (2010), “Precompetitive Research: A New Prescription for Drug
     Development”, Clinical Pharmacology & Therapeutics, 87, pp. 521-523.

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

                                             Chapter 6

                The importance of knowledge valuation for
                    knowledge networks and markets

      This chapter discusses future trends in the valuation of biotechnology and
      pharmaceutical assets as well as the companies that hold them, the
      development of new forms of financing health research, and how these
      developments impact on and may be influenced by knowledge networks and
      markets (KNMs) and new models of knowledge management. Several open
      questions remain, including: will improved valuation of biomedical firms and
      their assets also entail a better understanding and use of such intellectual
      assets and, if so, how might KNMs facilitate progress? And, can KNMs be
      constructed and operated so that they successfully reduce financial risk to
      their participants as well as increasing efficiency of innovation through
      collective action?


         Value creation in the biomedical sector will stall if the industry does not
     have access to sufficient sustainable, large-scale, long-term funding such
     that it can take medicines through the capital-intensive clinical development
     phases.1 Revenues from approved medicines ultimately drive value creation.
     The traditional equity funding model of life sciences is under pressure on a
     number of fronts (Welzl, 2008):
         • Regulatory and social demands to reduce the cost of drugs.
         •   Patent expirations and generic competition.
         •   Heightened pressure on budgets requiring increased efficiency.
         •   Declining R&D productivity.
         •   Restrictive accounting treatment of R&D costs.
         •   Absence of funding alternatives.
         •   Increased volatility in the equity markets.
         Arguably, public equity is not designed for the biotechnology industry
     because the public-equity system is set up to assess firms on the basis of
     earnings. This and other factors make the biomedicine sector extremely
     dependent on venture capital funding.
         But there are three significant problems facing life sciences innovators
     (Kimbrough, 2008). First, the average individual venture capital investment
     is USD 3 million, while the average total development cost of a pharma-
     ceutical product is around USD 1 billion. Second, there is a mismatch
     between the investment duration preferred by investors and typical develop-
     ment life cycles. Venture capitalists prefer to see returns within two to three
     years while the development time for drugs can be a decade or more. Third,
     the supply of venture capital has been demonstrated to be highly cyclical
     (Figure 6.1) and so quite volatile in the context of investments with long
     return times. To reduce the impact of this uncertainty, venture capitalists
     spread out their investments over a number of firms. This reduces their risks
     but limits the access any particular firm has to capital and thus creates a
     mismatch between investments and cost of development.
         This is particularly important given the ample evidence that resource
     allocation is sub-optimal in the life-sciences industry. The industry’s
     performance has been disappointing (Pisano, 2006), both in terms of the
     commercialisation of scientific breakthroughs by biotechnology firms and in
     terms of the ability of health-care providers to deliver high quality health
     care at low cost.

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
        6. THE IMPORTANCE OF KNOWLEDGE VALUATION FOR KNOWLEDGE NETWORKS AND MARKETS –                                       101
                                     Figure 6.1. Venture capital is highly cyclical

                                             United States                                  Europe

 Billions USD at current exchange rates
         1995     1996     1997     1998   1999   2000   2001   2002   2003   2004   2005     2006   2007   2008   2009   2010
Source: OECD (2011).

              The efficient functioning of markets should lead to resources flowing to
         their highest and best use (Kimbrough, 2008). It requires the availability of
         reliable information regarding an enterprise’s economic status, because
         stakeholders in the enterprise make decisions based on their perception of its
         economic status. Managers have incentives to manage the process of
         communicating with stakeholders regarding the enterprise’s economic status
         and are often also required to do so.
              The communication of information regarding a firm’s economic status is
         particularly important in the biotechnology and health industries for a
         number of reasons, including the inherent uncertainty of the industry.
         Scientific inquiry in bioscience is more uncertain than in other industries.
         The acceptance of scientific breakthroughs is uncertain due to regulatory
         roadblocks, there is a complex system of payments and reimbursements, and
         there are conflicting incentives of various actors. This uncertainty makes it
         difficult for potential investors to assess the expected returns on investments.
             Previous work (OECD, 2008) on the impact of R&D, patents, human
         capital and software shows that the average return on investment in
         intellectual assets can be large. Clear reporting of the intellectual capital
         held within a business is a theoretically sound way to help value an entity
         and thus raise investment.
                However, getting it “right” is proving to be a challenge.


                            Box 6.1. Case studies in biomedicine
    There are a number of cases whereby organisations have sought to provide
 supplementary information to stakeholders to improve investments and operations. The
 experience of the Karolinska Centre for Molecular Medicine (CMM) in Sweden, and
 HELIOS Kliniken, a hospital management entity in Germany, can both be seen as
 learning exercises.
    The Centre for Molecular Medicine (CMM) relies both on private donors and public
 funds. Its objectives for generating and distributing additional reporting were to promote
 cross-pollination among its resident scientists and to reduce duplication of effort, to track
 the use of capital and increase donor comfort with the organisation’s overall approach to
 scientific inquiry. In particular, the organisation hoped that the additional reporting
 would encourage donors to give more directly to the centre rather than tying donations to
 specific projects. Last but not least, the organisation sought to increase its scientific
    The dissemination of additional reporting appeared to be generally successful. It
 promoted the cross-pollination of research in the CMM and stimulated external media
 coverage raising the profile of the Centre; however, donors still wanted to have one-on-
 one contact with the CEO. Centre management has since adopted a “light” version of the
 additional reporting in the hope this would be more easily digestible by donors and thus
 have a larger impact on donations.
    HELIOS is a hospital management entity in Germany and, until recently, was
 independently publicly traded. The firm’s motivation for provision of additional reports
 was to become a more competitive provider as German hospitals became privatised, to
 promote Germany as a place to work, to achieve transparency with employees and to
 streamline the firm’s operations. The dissemination of additional reports led to structural
 changes in the firm. It was subsequently acquired by Fresenius, which decided HELIOS
 was a desirable acquisition in part through the additional detail in non-financial reports.
 However, other investors and analysts largely ignored the additional reports, perhaps
 because of overlaps with existing material. Nevertheless, HELIOS plans to continue with
 the additional reporting. Its next steps will be streamlining the reports, as well as making
 links between non-financial measures and financial outcomes.
    Both CMM’s and HELIOS’ use of additional reports were learning exercises. In both
 cases, the entities were able to achieve some of their objectives though with the
 important exception of linking non-financial reporting to the investment behaviour of
 many external stakeholders. These case studies suggest that current approaches to
 additional reporting are an effective tool for internal management and discussion, but
 that key success factors are hard to quantify and communicate to investors in a
 parsimonious and credible fashion. Current investor models are not designed to use non-
 financial information. It is difficult to use these additional types of information to
 identify winners and losers because this non-standardised information does not lend
 itself well to comparative decision-making. This is a challenge that will need to be
 addressed if additional reporting, including intellectual capital reporting, is to promote
 the more efficient use of capital in the health industry
 Source: Michael Kimbrough, 2008 OECD Workshop on Knowledge Networks and Markets.

                                       KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

Measuring and reporting intellectual assets at the firm level

           High performing firms in the life-sciences sector have several options
       for increasing their access to capital over more time by disclosing informa-
       tion indicating their high level of performance. Traditional accounting
       statements are insufficient because accounting does not recognise value
       creation in a timely manner in an industry with long development cycles.
       Some forms of additional reporting that are intended to lengthen the
       investor’s horizon beyond the typical two to three years include:
            •    Information on R&D efforts. Academic evidence suggests that
                 disclosure of the progress of a firm’s R&D portfolio is associated
                 with lower investor perceptions of risk.
            •    Supplemental disclosures. Some firms host conference calls with
                 shareholders when quarterly reports are issued. Others host R&D
                 open houses which are timed to mark passing Phase 1 or Phase 2
                 trials. These open houses offer access to scientists such that inves-
                 tors can assess research teams and the research team’s overall tech-
                 nological approach.
            •    Intellectual capital reports. There is early evidence that intel-
                 lectual capital reporting is a viable means of resolving the inherent
                 conflict between investors’ typical short investment horizons and
                 the long development cycles in the life sciences industry. Michael
                 Kimbrough’s research (Kimbrough, 2008) suggests that financial
                 information and non financial information are complementary and
                 that intellectual capital reporting seems to be effective at promoting
                 a more efficient allocation of resources in internal markets by pro-
                 moting cross disciplinary synergies. That said, it is an open question
                 whether or not intellectual capital reporting can fundamentally alter
                 capital flows in external capital markets. Investor models are not
                 oriented toward incorporating non financial information. There is a
                 danger that provision of this data will simply result in information
                 overload and to such measures being ignored completely by inves-
                 tors, regardless of their information content.
            Some firms employ intellectual asset (IA) evaluations as a routine part
       of their reporting. These evaluations are typically included as an element of
       annual reports. They can be presented as a separate intellectual capital report
       or provided within the management’s discussion and analysis (MDA) or
       management commentary. Skandia was the first firm to do so, and it has
       since been joined by others. For example, Infosys Technologies engages in
       corporate value reporting. For 15 years, they have integrated information on
       strategy and intellectual assets into their reporting. Among other items, their


     reports include an economic value-added statement, and they include IA on
     the balance sheet. Their analysis includes three types of IA: external
     structures (i.e. client relationships); internal structures (i.e. the firm’s R&D
     investments, organisational quality in terms of management processes); and,
     competence (i.e. the firm’s investment in human capital).
         Recent OECD work on corporate reporting highlights developments in
     guidelines and frameworks concerning intellectual assets (OECD, 2006 and
     2007) in two broad categories: i) narrative statements and non-financial
     reporting related to organisational performance; and ii) specific reporting
     about intellectual assets. Current practices often focus on backward-looking
     information, providing little systematic information about the capacity of the
     company to generate future revenues from intellectual assets.
         Similarly, there is so far little consensus regarding the nature of IA and
     the best approaches for their assessment. The valuation of IA involves many
     players with different views of the assets. The closest IA valuation comes to
     achieving common ground is acceptance of some definitions, especially the
     OECD definition of Intellectual Capital, which states that an intellectual
     asset is a “resource utilised in future value creation without a physical
     embodiment” and that types of these assets include proprietary knowledge,
     human capital, relational capital, and organisational capital. The Commission
     on Intellectual Capital (CIC) (Box 6.2) used a compatible definition that
     breaks down IA into: human capital: staff and management skills; software;
     R&D and innovation; brands and patents; strategies; processes; and relation-
     ships with suppliers and customers (Welzl, 2008).
         Nevertheless, there are indications that reporting on intellectual assets,
     done successfully, can bring efficiency and value creation benefits, improve
     the ability of firms to secure funding at a lower cost of capital, and better
     allocate resources (OECD, 2008). In particular, the adoption of intellectual
     asset reporting should contribute to mitigating the difficulties encountered
     by research-intensive SMEs to find financing for their research and innova-
     tion projects (Kimbrough, 2008).
          There is evidence that markets take into account the expected value of
     new innovations, R&D initiatives, technological breakthroughs and the quality
     of management using, for example, information provided by analysts and
     specialised sector publications (Darby et al., 1999) and directly in discussion
     with firms’ management. However, these ways of obtaining information about
     intellectual assets and business strategies implies additional costs and delays
     the dissemination of assessments in financial markets (Holland, 2002).

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

                  Box 6.2. Commission on Intellectual Capital Framework
     The Commission on Intellectual Capital (CIC) was created in Europe in 2006 and
  tasked with the creation of a framework for disclosure of non-financial information and
  the identification of intellectual capital (IC) key performance indicators (KPIs). The CIC
  released its first major report in 2008. The “Principles for Effective Communication of
  Intellectual Capital” provides a framework for IC reporting. The principles state that
  intellectual asset reporting should:
    •    Be standardised
    •    Make a clear link between valuation of intellectual assets and future value creation
    •    Be reliable, and assessments should be conducted consistently over time
    •    Be conducted responsibly, consistently and with transparency
    •    Balance disclosure and privacy
    •    Limit reporting to key metrics to prevent information overload
    •    Include risk assessments
    •   Disclose effectively, consider the placement and timing of reporting.
     The report also lays out the information that needs to be assessed and reported for the
  purposes of firm valuation. The CIC next plans to identify cross-sector differences,
  starting with the ICT industry.
  Source: Alexander Welzl, 2008 OECD Workshop on Knowledge Networks and Markets

            VC firms rely on their industry-specific human capital as their most
        valuable intellectual asset to identify good investment opportunities and to
        manage these investments (Gompers et al., 2005). However, access to
        information, including on firms’ intellectual assets, is crucial for VCs ability
        to enhance the value creation process.
             The next step in IA reporting will be linking IA evaluations to executive
        compensation models, and integrating them into the value-creation process
        at the level of employees (Kimbrough, 2008). These feedback systems and
        incentive structures should keep the firm focused on developing and making
        the best possible use of IA. The ultimate goal of the CIC is to create a
        virtuous cycle whereby companies focus more effort on knowledge assets,
        which increases efficiency of investments, which in turn supports more
        investment in knowledge assets, and so on (Welzl, 2008).

Exploiting intellectual assets to create national wealth

            Intellectual assets are not just of importance to firms, but to
        governments too. First, governments will need to ensure that they have
        mechanisms in place that support the moves to disclosure and reporting of
        intellectual assets in line with some of the emerging practices already


     discussed here. But second, a wholesale move to apply intellectual asset
     valuation to public intellectual assets in, say universities and public research
     organisations, could have unintended consequences if it is not got “right”. If
     got “wrong”, and intellectual assets are overvalued or poorly managed, there
     could be the potential for slowing down existing knowledge networks and/or
     public access to science. In a worst case scenario, valuing intellectual assets
     as part of an attempt to enter into monetised knowledge markets could
     become counter to properly functioning knowledge networks that aim to
     enhance knowledge circulation.
         The ability to create value from intellectual assets is contingent on the
     management capabilities of individual firms and the implementation of
     appropriate business strategies (OECD, 2008). Marr and Stratovic (2004)
     found that many companies declaring a desire to use knowledge manage-
     ment initiatives in order to create economic value from their intellectual
     assets did not have a clear idea of the exact expected benefits and required
     changes within corporate systems. The study argues that knowledge manage-
     ment is conceptually linked to organisational culture and processes, thus the
     overall target for companies should be to manage cultural and organisational
     means instead of knowledge.
         Lessons also might be drawn from analysis of the impact of policy
     relating to patenting of inventions in the public sector, For example, Bremer
     (2003) concludes that the three major things that contributed to the success
     of the Bayh-Dole Act in the United States were: certainty of title in the
     inventions; the inventor remains in the development picture (and so remains
     part of the team managing the knowledge asset); and there is uniformity in
     the handling of intellectual property under the law.
         But there is an extremely important third reason why governments
     should care about the valuation of intellectual assets – namely the very
     considerable shifts in the balance between government investments in
     intangible versus tangible assets in the United States (Figure 6.2) and else-
     where in the OECD zone.
         Given the quantitative importance of intellectual assets, their inclusion
     in measures of economic activity (such as GDP) is important for obtaining
     an accurate picture of economic growth, productivity and cyclical develop-
     ments (OECD, 2006). Corrado et al. (2006) argue that the conventionally
     measured capital stock is underestimated by some USD 1 trillion and the
     business capital stock by up to USD 3.6 trillion. Unsurprisingly, a number of
     governments are now putting significant effort into intellectual asset valua-
     tion (see, for example, Box 6.3) but so far there is little evidence that such
     efforts are focused on life sciences generally or biomedicine in particular.

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
   Figure 6.2. Shift in balance between investment in tangible/intangible assets in the
                                      United States

        Source: OECD, based on Corrado, Hulten and Sichel (2006).

               Box 6.3. Exploiting intellectual assets to create national wealth
     The Agency for Public Intangibles of France (APIE) was created in 2007 to develop
  an innovative and dynamic system of managing public intangible assets. Failure to take
  into account public intangible assets means losing out on a double dividend: direct
  financial benefits to government and benefits to the economy. Before the creation of
  APIE, many forms of intangible assets were undervalued and neglected because those
  managing the assets were not aware of their value and because they lacked the practical
  means to capitalise on the assets. APIE is currently working on methods to accurately
  value intangible assets in public accounting.
     APIE advises government agencies that have the assets under direct control. If the
  agencies collaborate with APIE, they receive an additional budget. Major APIE projects,
  targeted because they were most likely to be undervalued or underused, include
  databases and audio-visual assets as well as trademarks and know-how.
    APIE also has a role in projects that require horizontal collaboration within govern-
  ment, like defining IP rights in procurement contracts, arranging advertising on public
  websites, training government administration and creating the web portal.
  Source: Claude Rubinowicz, 2008 OECD Workshop on Knowledge Networks and Markets.


Leveraging knowledge to create equity markets

          Earlier in this chapter the suggestion was made that public equity markets
     might not be very effective for biotechnology companies. Of course, this is at
     best a generalisation that assumes all biotechnology companies are young and
     are yet to generate net profits – which clearly is not the case.2 Nevertheless,
     given the high risks and long lead times to market, getting access to public
     equity markets has proved challenging for a number of firms in the bio-
     medicine sector.
          Thus the pressures on the traditional equity model of life-sciences
     funding have encouraged pharmaceutical firms and others actively to seek
     alternative funding sources. Past experiments with alternative funding
     suggest that any solution must address several concerns (Brown, 2008). It
     must provide for risk transfer and diversification, ideally across different
     pharmaceutical firms, such that the solution offers industry access to the
     distribution of portfolio development risk and provides a reliable stream of
     low-cost developmental capital. Any alternative funding model should
     involve significant management involvement and scrutiny by independent
     parties and thus requires buy-in both from R&D and finance functions
     within the firm, and shareholders outside the firm.
          Pooling projects and spreading risk allows capital markets to formulate
     structured finance to fund developers, large and small, in developed and
     emerging economies. Such development finance facilities can – at least in
     principle – become self-sustaining, supported by and supporting knowledge
     markets. However, there are to date very few such experiments (or at least
     publicly acknowledged experiments) to draw upon in order to learn lessons.
     Box 6.4 describes one such experiment – SecurePharma – which aims “to
     promote and execute solutions for pharmaceutical companies, bringing
     development risk transfer and accessing a new source of stable, long-term
     funding….[based on]… structured methods to provide a new source of risk
     capital and liquidity from capital markets….for large portfolios of around 30
     projects in a range of therapies”.3
          It is to be hoped that better valuation of intellectual assets and a clearer
     acceptance by the market of IA reporting should allow more rational choices
     to be made about equity investment in the future and so contribute to much
     improved risk management. If this aspiration were to come about, it is not
     beyond the realms of imagination to see some kind of secondary market
     develop from such investments, which could have significant impacts on
     productivity and efficiency of investment. Governments could benefit from
     analysing how they might kick-start such secondary markets in certain high
     need areas. The OECD has already looked at some of the implications of
     applying knowledge networks and markets thinking in the area of neglected
     and emerging infectious diseases (OECD, 2009).

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

                     Box 6.4. Structured finance facility – SecurePharma
   SecurePharma is working to implement a Structured Finance Facility for Clinical
Development. In this model, biotechnology and pharmaceutical firms provide a pool of
compounds of interest. Compounds are selected by an independent scientific panel, and the
contributing firms share a stake in the development of the entire pool of selected compounds.
The development costs for the clinical-development phase are acquired through a bond issue,
which is made possible by the pooling of the technologies. The development of a portfolio of
technologies has less risk than the development of a single technology. Furthermore, pooling
the technologies brings the scale of the individual investment up to the minimal level that
would be considered in these capital markets (i.e. at least USD 250 million). In the unlikely
event that an entire pool fails, the costs are shared across bond investors, insurers, and the
firms that provide the candidate technologies. Thus, the investments are lower-risk than if the
whole cost were borne by the biotechnology or pharmaceutical firms. These firms get access
to low-cost funding and reduced risk in exchange for provision of compounds at sunk cost.
The development work is done by the firms that provide the compounds, or by contracted
firms. The firms doing the development are paid to do the work and are overseen by the
independent scientific panel.
   The SecurePharma funding structure should be able to increase the funding available for
development of therapies and thus increase the number of drugs in the development pipeline.
It may also provide a mechanism for funding clinical research for therapies that target NTDs.
To be successfully implemented, these kinds of structures require further research and policy
support. For example, the pooled clinical-development model will only function effectively,
in the long term, if methods are developed that allow the pool managers to account for
contributions of drug candidates.
   SecurePharma is making progress in realising this model of pharmaceutical development
funding. In one ongoing project, a commercial institution is in discussion about the purchase
of the risk end of bonds associated with mainstream pharmaceutical development, and a
number of small biotechnology firms are highly interested in participating in that transaction.
In terms of the funding of NTDs, there are portfolios approaching the registration phase
thanks to the Bill & Melinda Gates Foundation and other donors; however, these pools still
need further investments of capital. The United Kingdom government is involved in funding
and providing securitisation for certain vaccine projects. These kinds of projects may be able
to make linkages with the emerging markets opening up in developing countries and employ
securitisation models in support of emerging economies. An open question for further
consideration is: what will be the impact on future KNMs from the additional supply and
demand, and special needs of emerging economies?
Source: Peter Brown, 2008 OECD Workshop on Knowledge Networks and Markets.



         The recognition that investment in intellectual (intangible) assets now
     dominates investments in tangible assets and helps create value and drive
     innovation is longstanding, particularly since the work of Corrado et al.
     (2006). However, valuation of intellectual asserts still is in its relative
     infancy, with little agreement over standardisation or sector-sector (let alone
     firm-firm) comparability.
          In principle, the accurate valuation of intellectual assets should be a sine
     qua non for the biotechnology sector, so much of the value of firms clearly
     resting on know-how and intellectual property, and as Kimbrough argues
     markets should be more efficient if investors can reward firms that own such
     assets and can turn them into effective innovation and return on investment.
     However, case study work suggests that the potential advantages of accurate
     valuation of intellectual assets is not being realised because IA reporting is
     neither standardised, nor widely adopted, though it is arguably more widely
     accepted and utilised by market analysts than is commonly perceived by
     managers in the sector.
         Enhanced approaches to reporting can raise awareness – among
     businesses, investors and policy makers – of the potential for small firms to
     develop and exploit intellectual assets. And government could assist efforts
     to promote identification and dissemination of best practices in reporting.
     Dissemination of knowledge about the potential benefits could also
     encourage more companies to improve their disclosure practices as well as
     their internal management systems. Better information on intellectual assets
     in national accounts and in corporate reporting would also facilitate the
     design, monitoring and implementation of more efficient public policies, for
     example with respect to investment in intellectual assets to generate
     economic value (OECD, 2008).
         There are thus several reasons to presume that IA reporting will
     continue to grow. Firstly, IA reporting offers immediate benefits to firms by
     providing internal management with information required to make better
     decisions – The European Federation of Financial Analysts Societies’
     Commission on Intellectual Capital (EFFAS-CIC)’s4 so-called “virtuous
     cycle” of intellectual asset management. Secondly, there are on-going
     efforts to standardise IA. Amongst others, the EFFAS-CIC is working to
     promote IA reporting and to standardise its reporting methodology. Thirdly,
     the growth of well managed KNMs could reduce transaction costs and
     increase the fluidity of transactions involving intellectual assets – perhaps
     imparting demand-side pull. Better awareness and transparency related to
     the value of intellectual assets clearly needs to be part of future efforts, and

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

       the APIE in France seems to be proving an interesting model to watch in this
            The concept of KNMs exerting demand-side pull is well illustrated by
       the SecurePharma experience. The model is innovative in that it seeks to
       mitigate and distribute the risk of technological or market (even system)
       failure. To work well a robust system for IA evaluation would be helpful,
       since the long-term success of this approach requires a fair and objective
       assessment of the value of candidate therapies provide by pharmaceutical
       firms. This would be particularly true if a robust secondary market were to
       be developed to help maintain investment in relatively early stage risks.
           Monetised knowledge markets might help spread risk and bring
       otherwise underutilised inventions through to the market – but better means
       to evaluate, report and trade intellectual assets need to be developed to make
       for real dynamism. Indeed, the better understanding of the value of
       knowledge is at the core of a vibrant system of KNMs, irrespective of
       whether the knowledge and intellectual assets traded are monetised or not.

1.      This chapter draws on contributions by Alexander Welzl, Claude Rubinowicz,
        Peter Brown and Michael Kimbrough at the October 2008 OECD Workshop
        on Knowledge Networks and Markets, as well as on a range of other sources.

2.      Though in the view of Pisano (2006), a high profile commentator, the sector
        has significantly underperformed as a whole.

3.      SecurePharma website,,
        last accessed on 12 December 2011.

4.      Welzl (2008). See also


Bremer, H (2003), The Bayh-Dole Act: Impact on University Research and
     Intellectual Property Ownership Rights, Rensselaer Polytechnic Institute,
Brown, P. (2008), “Establishing a Clinical Trials Finance Facility: Finance and
     Licensing Proposal”, OECD Expert Workshop on Knowledge Markets in
     Life Sciences.
Cardullo, M. (1999), Technological Entrepreneurship: Enterprise Formation,
      Financing and Growth, Research Studies, Press Ltd., United Kingdom.
Corrado, C., C. Hulten, & D. Sichel, (2006), “Intangible Capital and Economic
      Growth,” National Bureau of Economic Research Working Paper 11948.
Darby, M., Q. Liu & L. Zucker (1999), “Stakes and Stars: The Effect of
      Intellectual Human Capital on the Level and Variability of High
      Technology Firms’ Market Values”, NBER Working Paper, 7201.
Gompers, P., A. Kovner, J. Lerner & D. Scharfstein (2005), “Venture Capital
    Investment Cycles: The Impact of Public Markets”, NBER Working Paper,
Holland, J. (2002), “Fund Management, Intellectual Capital, Intangibles and
      Private Disclosure” Department of Accounting & Finance, University of
      Glasgow, Working Paper, 2002/4.
Kimbrough, M (2008), “Financial Communication in the Life Sciences: Focus on
     Intellectual Capital Reporting”, OECD Expert Workshop on Knowledge
     Markets in Life Sciences.
Marr, B. and D. Stratovic (2004), “Understanding Corporate Value: Managing and
      Reporting Intellectual Capital”, School of Management, Cranfield
OECD (2006), “Creating Value from Intellectual Assets”, Meeting of the OECD
    Council at Ministerial Level”, OECD, Paris.
OECD (2006), “Intellectual Assets and Value Creation, Implications for Corporate
    Reporting”, OECD, Paris.
OECD (2007), “Intellectual Assets and Corporate Reporting: The Situation of
    Small Caps,” OECD, Paris.
OECD (2008), Intellectual Assets and Value Creation Synthesis Report, OECD,

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

OECD (2009), Noordwijk Medicines Agenda, Paris, and associated
    documentation. See
OECD (2011), OECD Science, Technology and Industry Scoreboard 2011, OECD
    Publishing. doi: 10.1787/sti_scoreboard-2011-en.
Pisano, G.P. (2006), Science Business: the Promise, the Reality and the Future of
      Biotech, Harvard Business School Press, Boston, Massachusetts.
Welzl, A. (2008) “Measurement, Reporting and Valuation of Intellectual Assets –
      The Investor View”, OECD Expert Workshop on Knowledge Markets in
      Life Sciences.


                                             Chapter 7

      Conclusions and research needs in knowledge networks
                          and markets

      This final chapter presents conclusions and policy recommendations. It
      identifies what are the common features of knowledge networks and markets
      (KNMs) and on what criteria they differ. It reiterates why KNMs matter for
      innovators in the biopharmaceutical sector, in particular because of their
      contribution to: addressing financial pressures, accelerating science and
      development, improving health outcomes, and improving the regulatory
      dialogue. The chapter concludes with a review of the support that KNMs
      currently enjoy from government and sets out proposals for possible future


What are knowledge networks and markets (KNMs)?

         The term “knowledge network and market (KNMs)” refers to the broad
     set of initiatives that seek to increase access to data and knowledge from
     widely distributed sources in order to facilitate further innovation. KNMs
     are of particular interest in the life sciences where, traditionally, much
     biomedical knowledge has been kept siloed or proprietary.
         This report seeks to identify common features across the range of
     different institutional arrangements we are calling KNMs. At a minimum,
     KNMs must be:
         •   Formal: Most of the KNMs currently in existence are governed by
             formal rules of engagement, many of which are embodied in legal
             documents (e.g. organisational charter and policies, partnership
             agreements). Therefore, creators of KNMs almost always need to
             address a number of governance issues, chiefly IP rights, data-
             access policies and privacy restrictions, as well as addressing the use
             of human subjects. OECD instruments (OECD, 2002; 2009) may be
             amongst the sources KNMs managers might turn to.
         •   Transformative: KNMs are created to reorganise resources and
             facilitate inter-organisational relationships, and in so doing they
             transform the “knowledge collective”. Often the reorganisation of
             resources increases the value of those resources. For example, a
             pool of biomarkers that have been expertly chosen and validated is
             worth considerably more than the candidate biomarkers are worth
             when they are distributed across a number of firms and not yet
             validated. In other cases, the reorganisation of resources permits a
             freer flow of information. For example, caBIG links the data
             produced by clinical care, clinical research and scientific discovery
             to synergistically improve all three areas of endeavour. This
             transformative quality appears to be desirable when organisations
             are faced with problems that can only be addressed through
             collective action (e.g. biomarkers consortia) and problems that
             cannot be addressed through existing structures (e.g. developing
             therapies for NTDs). In some cases, the transformation appears to be
             effected by reducing transaction costs, thus providing access to
             resources unfeasible to access without the KNM. For example,
             InnoCentive created access to its solver community by providing the
             formal structure linking answer seekers and solvers.

                                   KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

            •    Translational: KNMs are developed to facilitate the further
                 development of data, information and knowledge; most life-sciences
                 KNMs’ ultimate impact will be to increase the quality of medical
                 care through improved therapies, etc. On-line references that do not
                 allow users to build upon the data provided would probably not be
                 classified as KNMs.
            Knowledge networks and markets do not appear to be defined by:
            •    Knowledge type: KNMs appear to exchange many intellectual
                 assets including bio-specimens and other materials, services, data-
                 bases, expertise, software, methodologies and IP, including patents.
                 The types of knowledge exchanged are not limited to pre-compe-
                 titive assets (e.g. InnoCentive).
            •    Structure: KNMs are structured in a wide variety of ways,
                 including public-private partnerships (e.g. TI Pharma’s initiatives),
                 consortia (e.g. The Biomarkers Consortium), prize mechanisms
                 (e.g. InnoCentive) and data-sharing platforms (e.g. caBIG).
            •    Financing models: The initiatives have a variety of financing
                 models, including “pay to play”, public funding, advertising, fees
                 for service, and some combination of the preceding items. Indivi-
                 dual exchanges may or may not be monetised.
          Regardless of the details of a definition, it is clear that knowledge
       markets are a real and growing presence in the life sciences and beyond.

What benefits might knowledge networks and markets offer?

           KNMs can have benefits in four related areas: i) addressing pressures on
       the biotechnology and pharmaceutical industry; ii) improving health
       outcomes; iii) accelerating scientific progress; and iv) regulatory-industry

       Addressing pressures on the biotechnology and pharmaceutical
            The biotechnology and pharmaceutical industry is challenged on a number
       of fronts, including:
            •    Rising drug-development costs.
            •    Shrinking numbers of drugs in the development pipeline.
            •    Regulatory and social demand for cheaper drugs.


         •   Concern about the openness of clinical trial data and suppression of
             negative results.
         •   Patent expirations and generic competition.
          There is a real concern that the conventional approach to drug develop-
     ment is unsustainable, and this is reflected in pharmaceutical company
     reports (GSK, 2010; Pfizer, 2010). This pressure will heighten as the life
     sciences move towards a personalised medicine paradigm. Personalised
     medicine is likely to generate moderately disruptive technologies. Drugs
     will have to be developed and marketed with tests that predict the drug’s
     efficacy and/or monitor its toxicity. The introduction of the tests will have
     implications for relationships between drug providers, health professionals,
     patients and payers. Personalised medicine is also very likely to result in
     reduced market sizes for individual therapies, and the need for tests may
     increase the cost of developing therapies and getting them approved for sale.
     These changes will increase the rate at which the current approach to drug
     development fails unless steps are taken to increase the industry’s efficiency
     and the volume of therapies under development.
         In response to these pressures, firms are seeking new ways to operate
     and are open to engaging in KNMs. Firms participate in KNMs for many
     reasons, including reducing their research costs, sharing risk, gaining access
     to necessary resources (e.g. compound libraries), and participating in the
     development or revision of regulatory approval processes. KNMs also allow
     firms to experiment with new research strategies and business models,
     including, for example, InnoCentive’s “hive” business model and caBIG’s
     research model for integrating science, clinical care and clinical trials.
     Through such experiments, the industry may be able to speed up the rate at
     which it adapts to its changing environment and find new models that will
     allow it to be sustainable in an era of personalised medicine.
         Since KNMs are mechanisms that support the biotechnology and
     pharmaceutical sector, countries have a strong economic interest in promoting
     KNMs among domestic firms. It may be appropriate to link supports for
     KNMs to policies that support high technology clusters.

     Improving health outcomes
          The competitiveness of the biotechnology and pharmaceutical industry
     has important implications for health-care systems and health-care delivery.
     Addressing productivity decline, incentivising research in areas where there
     is a public-health need and reducing health innovation development costs
     will help OECD countries better meet the health-care needs of their popula-
     tions. Together with the transition to personalised medicine, improvements
     to the industry could yield:

                                  KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

            •    Growth in evidence-based treatment options.
            •    More transparency of health information.
            •    Early adoption of treatments.
           Advances in R&D approaches and experiments with new models for
       research and new business models should also have an impact on global
       health priorities. For example, novel clinical-trial funding models may
       stimulate work on neglected diseases (OECD, 2009).

       Accelerating science and development
           Knowledge exchange in scientific enterprise needs to be modernised in
       order to keep up with accelerating data generation. Knowledge networks and
       markets should facilitate international collaboration such that discovery is
       powered by a global research community, and encourage interdisciplinary
       work by breaking down the barriers to collaboration. Together, these
       advances should advance the rate of scientific discovery by bringing new
       approaches, disciplines and modes of work to bear on problems. KNMs can
       also increase access to data and thus facilitate the mining of existing data,
       and help to improve clinical research and practice by closing the feedback
       loop between patients, clinicians and researchers.

       Improving regulatory-industry dialogue
            Recent industrial changes and scientific advances have the potential to
       lead to new treatment paradigms which challenge our current regulatory
       approaches to health-care products and processes. KNMs facilitate inter-
       institutional dialogue and can help build agreement on methodologies,
       standards and outcomes. Such dialogue is critical to create a predictable,
       transparent and trusted regulatory system that encourages needed innova-
       tions for health. Recent examples include:
            •    Regulation of personalised medicines and targeted therapies.
            •    Biomarker validation and the design of regulatory approvals process
                 for biomarkers.

What policy support do KNMs require?

           OECD Working Party on Biotechnology analysis identified four areas in
       which Knowledge networks and markets may require support: i) ICT
       infrastructure development and maintenance; ii) IP rights; iii) IA valuation;
       and iv) regulation and legislation.


     ICT infrastructure development and maintenance
         In order to achieve their objectives, KNMs usually require sophisticated
     ICT infrastructures. There are currently two orthogonal paradigms for KNMs
     systems: containers and networks. Both permit geographically distributed
     access to knowledge, manage permissions and are capable of supporting the
     exchange of a wide range of types of data. Both types of ICT infrastructure
     require long-term support in the form of continued funding for the creation
     and maintenance of the systems. It is important that these designs be
     “technology-neutral” so that they can grow and adapt as ICTs evolve.
     Governments need to provide such infrastructure support, especially when the
     KNMs will also disseminate valuable government data (e.g. NLM’s initia-
     tives), if optimal productivity gains are to be achieved.

     Intellectual property (IP) rights
         IP rights are a key consideration in KNMs because they help protect
     both the participants’ and the KNMs’ intellectual assets. The governance
     and management of KNMs will need to take clear and explicit decisions
     regarding IP as these will shape participant expectations. There is a concern,
     however, that agreeing the terms for IP protection and exploitation could be
     time-consuming and costly when establishing KNMs and so potentially
     present barriers. Governments could support the creation of KNMs by
     developing general IP frameworks and guidelines for them (the United
     Kingdom Lambeth agreements (OECD, 2011), for example, have been quite
         The consensus is that publicly funded research organisations are the
     least amenable to negotiating broad access to their IP rights. Governments
     should consider providing IP guidance to the research organisations that
     they fund regarding access to publicly funded research.1

     Intellectual-asset (IA) valuation
          Intellectual assets are notoriously difficult to assess in a transparent and
     reliable manner. Work is underway to standardise IA valuation, and firms
     are finding value in conducting valuations even when simply as an internal
     management tool. Some KNMs would benefit from an accepted and robust
     methodology for IA valuations. These two areas of practice will have to co-

                                    KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

       Regulation and legislation
           Regulations and legislation affect the incentives for participation in
       KNMs. There is a need to reduce bureaucracy and red tape, and to ensure
       that conflict-of-interest rules and antitrust regulations are appropriate.

What should be the focus of future work?

            Many are optimistic about the proliferation of KNMs and their potential
       impacts. It is clear that governments must make use of KNMs to advance
       their health innovation and health-care goals. Policies can help in a variety
       of ways, including by enabling and catalysing partnerships, providing long-
       term continuous funding for KNMs infrastructure, contributing valuable
       data and providing resources that reduce the administrative or legal burden
       of creating KNMs (e.g. guidelines for IP agreements). As a rapidly emerging
       trend in the life sciences, KNMs appear to be well-designed to help this
       industry through a major transition. They seem to have the potential to
       deliver significant socio-economic benefits. However, the extent to which
       this emerging trend and its benefits are broadly realised will have to be the
       subject of on-going monitoring and study.
            Many questions remain about how to characterise the current trend and
       how to draw lessons for the broader life science research community. There
       is likely to be much debate on the definition of KNMs given the variety of
       different understandings of what “openness” actually means or strives to
       achieve. The proliferation of different initiatives that offer boundless
       permutations on their structure, function and membership means that little
       can be said about KNMs with great certainty and a more restrictive
       taxonomy of organisational types might be necessary to draw any
       meaningful best practices. These advances need to be made without undue
       sacrifice to academic introspection. The main issue is that governments and
       practitioners find ways to make KNMs work as efficiently and as effectively
       as possible.

Future research programme

            Future work that the OECD is considering taking up include:
            •    Transforming large-scale shared infrastructures platforms for life
                 sciences via next generation KNM:
                     The development of policy principles for integrating complex,
                     high-value data and new IT platforms to enable the creation of
                     new tools in biotechnology;
                     Convergence of new biology, IT, physical sciences.


         •   Enabling social networking in life sciences:
                 Development of a policy brief about the impact of social
                 networks on health innovation.
         These two themes represent complementary emerging issues that
     increasingly drive health innovation. On the one hand, there is a need to
     understand how to derive greater value from advances in the life sciences in
     large-scale government, and institution- and industry-backed efforts that go
     beyond traditional PPP. The question for governments is how to create
     large-scale interdisciplinary networks that reach beyond the simplistic focus
     on either producing publications or delivering profits. These networks can
     and must do both, but this makes their organisation and governance less
     obvious. On the other hand, social networks, made up of individuals, are
     having a growing impact on research, clinical applications, patient choices
     and care. The two trends are new, happening simultaneously and using
     KNMs to achieve their goals.
         •   Next-generation networking of research infrastructures. IT allows
             the networking of disparate life science databases and repositories
             and opens up enormous research and business opportunities. While
             the ability to develop, access and use large-scale databases and
             resources from multiple disciplines and locations has opened new
             avenues of research and new fields, it has also made information
             mining and knowledge integration increasingly complex. Research
             infrastructures in the life sciences need to be sustainably funded,
             they need to be technology-neutral so that systems are adaptable and
             do not limit the future scope of research or collaborations, and they
             need to be accessible, possibly at different costs for different
             purposes. It would be worthwhile to explore where and how the
             networking of research infrastructures is happening, for what
             purpose, how it is being funded, what technological neutrality in
             practice it entails, what are the challenges and to what extent this
             networking leads to market creation for knowledge/data.
         •   The networking of biomedical data. This is of particular interest for
             its potential contribution to the delivery of personalised medicine
             through both broad population analyses and targeted personal health
             information. The combination of health and biological data sources
             is occurring in public, private and mixed initiatives. Examples of
             networks of biomedical knowledge and data include CaBIG and the
             BC Cancer Agency, as well as private endeavours such as Google
             Health, Microsoft Health, IBM, Fujitsu Health, Kaiser Permanente
             and many others. These networks can increase the evidence base
             available for research as well as improve health-care efficiency by

                                  KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012

                 providing more information to clinicians. The line separating know-
                 ledge networks and knowledge markets is hard to draw. The role of
                 governments in initiating or participating in such endeavours is
                 important to understand.
           OECD countries have a strong policy interest in knowledge networks
       and markets as a tool to achieve many different economic and social goals in
       health and even more broadly in the life sciences. While the examples of
       active KNMs described in this report are very promising, more information
       is needed to determine how KNMs can best be employed to achieve specific
       policy objectives.


1.    The work of the Royal Society on science as a public good is notable in this regard.


GlaxoSmithKline (GSK) Annual Report (2010),
OECD (2002), “Genetic Inventions, Intellectual Property Rights and Licensing
    Practices: Evidence and Policy”, OECD, Paris.
OECD (2009), The Bioeconomy to 2030; Designing a Policy Agenda, OECD,
OECD (2011), “Collaborative Mechanisms for Intellectual Property Management
    in the Life Sciences”, OECD, Paris.
Pfizer Annual Review (2010),

                                  KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                    BIBLIOGRAPHY    – 125


Aghion, P., P. Howitt, and G. Violante (2000), “Technology, Knowledge and
     Within-Group Inequality”, University College of London.
Apostolou, D. et. al. (2005), “Challenges and Directions in Knowledge Asset
      Trading” International Journal of Intelligent Systems in Accounting,
      Finance & Management, Vol. 13, No. 1, March 2005, pp. 1-15(15).
Arora, A., A. Forsfuri, A. Garmbardella (2001), “Markets for Technology and
      Their Implications for Corporate Strategy,” Industrial and Corporate
      Change, Vol. 10, No. 2, pp. 419-51.
Bingham, A. (2006), “Whither Pharma,” Drug Discovery World Business,
      pp. 9-15.
BioVentures for Global Health (2007), Closing the Health Innovation Gap: A role
     for the Biotechnology Industry in drug discovery for neglected diseases.
     Available online at:
Buetow, K. (2008), “Heading for the BIG Time: The NCI's bioinformatics
     network, caBIG, integrates cancer data from all over the US,” The Scientist,
     22:4, pp. 60-66.
Callan, B., and I. Gillespie (2007), “The Path to New Medicines,” Nature, 449,
      pp. 164-65.
Cohen, W., et al. (1997), “Appropriability Conditions and Why Firms Patent and
     Why They Do Not in the American Manufacturing Sector,” Working Paper
     (Pittsburgh: Carnegie-Mellon University).
Gassman, O. and G. Reepmeyer (2005), “Organizing Pharmaceutical Innovation:
     from Science-based Knowledge Creators to Drug-Oriented Knowledge
     Brokers.” Organizing Pharmaceutical Innovation, 14:3, pp. 233-45.
Henry, G., (2002), “Patents, Innovation and Access to New Pharmaceuticals,”
      Journal of International Economic Law, 5(4), pp.: 849-60.


Jarboe, K.P. (2008), “Building a Capital Market for Intangibles,” Intellectual
      Asset Management, June/July, pp. 30-35. Available online at:
Lettl, C. (2005), “The emergence of radically new health care technologies:
       Inventive users as innovation networkers” Technology and Health Care, 13,
       pp. 169-83.
Maass, W., W. Behrendt, and A. Gangemi (2007), “Trading digital information
      goods based on semantic technologies” Journal of Theoretical and Applied
      Electronic Commerce Research, Vol. 2 Issue 3, December, pp.: 18-35.
OECD (2002), Benchmarking Industry-Science Relationships, OECD, Paris.
OECD (2003), Turning Science into Business: Patenting and Licensing at Public
    Research Organisations. OECD, Paris.
OECD (2005), Recommendation of the Council on the Licensing of Genetic
    Inventions, OECD, Paris.
OECD (2006). “Research Use of Patented Knowledge: A Review”, OECD
    Science, Technology and Industry Working Paper 2006/2,
OECD (2006), Valuation and Exploitation of Intellectual Property. OECD
    Science, Technology and Industry Working Paper 2006/5, OECD, Paris.
    Available online at:
OECD (2006), “Intellectual Assets and Value Creation: Implications for Corporate
    Reporting”, OECD, Paris. Available online at:
OECD (2006), “Principles and Guidelines for Access to Research Data from
    Public Funding”. OECD, Paris. Available online at:,3343,en_2649_34293_38500791_1_1_1_374
OECD (2007), “Policy Brief: Creating Value from Intellectual Assets. OECD”,
    OECD, Paris. Available online at
OECD (2007), Participative Web and User-Created Content: Web 2.0, Wikis and
    Social Networking, OECD Publishing, Paris.
OECD (2008), OECD Recommendation on Enhanced Access to Public Sector
    Information. OECD, Paris. Available online at:

                                  KNOWLEDGE NETWORKS AND MARKETS IN THE LIFE SCIENCES – © OECD 2012
                                                                    BIBLIOGRAPHY    – 127

OECD (2008), The Internationalisation of Business R&D: Evidence, Impacts and
    Implications, OECD Publishing, Paris.
OECD (2008), OECD Information Technology Outlook 2008, Chapter 3 on “ICT
    R&D and Innovation”, OECD Publishing, Paris.
OECD (2008), “Innovation within Global Value Chains – The Emergence of
    Global Markets for Technology: Issues Paper”,
OECD (2008), “Survey of Patent Licensing,” OECD, Paris.
OECD (2008), “The Innovation Strategy: ICTs and Innovation,”
OECD (2008), “The Seoul Declaration for the Future of the Internet Economy.”
    Available online at:
OECD (2008), OECD Policy Guidance on Digital Content.
OECD (2008). “OECD Recommendation on Human Biobanks and Genetic
    Research Databases”,
OECD (2009), “Bioeconomy 2030, Designing a Policy Agenda”. Paris, OECD
OECD (2009), Pharmacogenetics: Opportunities and Challenges for Health
    Innovation, OECD Publishing, Paris.
OECD, (2011). “Knowledge Markets in the Life Sciences: Issues Paper”,,3746,en_2649_34537_39407351_1_1_1_1,0
Pharma Futures (2007). Prescription for Longterm Value. Available online at:
Shane, S. (2001), “Technology Regimes and Firm Formation,” Management
      Science, 47(9): 1173-90.
Spear, B., M. Heath-Chiozzi and J. Huff (2001), Trends in Molecular Medicine,
       Vol. 7, pp. 201-04.
Teece, D. (1998), “Capturing Value from Knowledge Assets: The New Economy,
      Markets for Know-How, and Intangible Assets,” California Management
      Review, 40:3, pp. 55-79.
Wang, S.S. (2008), “Genetics may bring new life to failed drugs,” The Wall Street
     Journal, 24 March. Available online at:

                     AND DEVELOPMENT
     The OECD is a unique forum where governments work together to address the
economic, social and environmental challenges of globalisation. The OECD is also at the
forefront of efforts to understand and to help governments respond to new developments
and concerns, such as corporate governance, the information economy and the challenges of
an ageing population. The Organisation provides a setting where governments can compare
policy experiences, seek answers to common problems, identify good practice and work to
co-ordinate domestic and international policies.
     The OECD member countries are: Australia, Austria, Belgium, Canada, Chile, the
Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland,
Israel, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland,
Portugal, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, the United Kingdom
and the United States. The European Union takes part in the work of the OECD.
     OECD Publishing disseminates widely the results of the Organisation’s statistics gathering
and research on economic, social and environmental issues, as well as the conventions,
guidelines and standards agreed by its members.

                        OECD PUBLISHING, 2, rue André-Pascal, 75775 PARIS CEDEX 16
                          (93 2012 01 1 P) ISBN 978-92-64-11937-6 – No. 60125 2012
Knowledge Networks and Markets
in the Life Sciences
Chapter 1. The rise of knowledge networks and markets as enablers of open innovation
Chapter 2. Knowledge flows
Chapter 3. Advantages of knowledge networks and markets
Chapter 4. Theories for building knowledge networks and markets
Chapter 5. Case studies of knowledge networks and markets
Chapter 6. The importance of knowledge valuation for knowledge networks and markets
Chapter 7. Conclusions and research needs in knowledge networks and markets

  Please cite this publication as:
  OECD (2012), Knowledge Networks and Markets in the Life Science, OECD Publishing.
  This work is published on the OECD iLibrary, which gathers all OECD books, periodicals and
  statistical databases. Visit, and do not hesitate to contact us for more

                                                 ISBN 978-92-64-11937-6
                                                          93 2012 01 1 P      -:HSTCQE=VV^X\[:

To top