1995 ISSN 1756-8919 Future Med. Chem. (2011) 3(16), 1995–2020 10.4155/FMC.11.160 © 2011 Future Science Ltd PersPective For the last few decades small-molecule drug discovery has been dominated by a linear reverse chemical genetics (or reductionist) approach, in which drug discovery was considered to be a modular, sequential, and scalable process that could be readily industrialized [1]. This approach has led to a greater overall emphasis on business rather than science in the pharma- ceutical industry but, unfortunately, has not provided long-term financial sustainability [2]. Outsourcing has long been an important part of clinical and development activities but recently more emphasis has been placed on augment- ing in-house research via outsourcing to CROs, specializing in both discovery and preclinical activities. This trend, adopted by both phar- maceutical and biotechnology companies, offers distinct advantages in terms of both flexibil- ity and access to knowledge and expertise, but alone does not necessarily address the core problems of the pharmaceutical industry, such as lack of innovation or the lack of new chemi- cal entities. The pharmaceutical industry has always worked with third parties to access spe- cific assets, technology, and expertise. Initially, this approach remained constrained within the reductionist framework and was largely a method of reducing fixed operating costs. More recently, the continual development of CRO capabilities, improvements in data sharing, and the development of long-term interactions between CROs and clients are all having a sig- nificant impact on strategic thinking through- out pharmaceutical companies. However, to ensure long-term success for this industry, there is the need to embrace new types of R&D part- nerships to access and capitalize on knowledge in order to create and maintain new sources of innovation [3]. An alternative to contracting out of research is the construction of networks between centers of excellence, often in an open approach, which can provide a flexible and potentially more adapt- able environment. In essence, collaboration and outsourcing strategies are, thus, redefining the pharmaceutical landscape. Generally, the more innovative the target or medicine, the greater the need for collaboration and knowledge sharing between experts, so to revise the over- all drug-discovery paradigm we need to better bring together expertise across research centers. Individual centers would be able to concentrate on developing their respective areas of exper- tise but simultaneously capitalize on the exper- tise of the other centers within a collaborative network. In this context, collaboration versus outsourcing are dissimilar and context-depen- dent alternatives but are not mutually exclusive options. Both have significant areas of overlap and both have considerable potential for further development. Overall, they offer significantly different approaches to addressing the innova- tion and knowledge gaps in drug discovery. In particular, they offer different opportunities for addressing the need for improvements in the clinical/research interface and more predict- able translation of successful compounds into clinical studies. Collaboration versus outsourcing: the need to think outside the box As has been widely reviewed elsewhere, the pharmaceutical industry is experiencing an ‘innovation deficit’ as evidenced by the decline in new chemical entity output. This decline, compounded by increased costs and regulatory requirements highlights the need to significantly revise strategic options across the drug-discovery spectrum. Within such revision(s), much of the focus has been on outsourcing to reduce, or at least contain, costs, but if the underlying predominance of ‘closed collaborations’ is not challenged to allow better use of combined knowledge and, thus, move towards a more genuine collaborative process then a ‘numbers only’ approach will not bring medium-to- long-term survival. There are many problems to confront in evolving new sustainable strategies, a real need to think differently exists and should to be cultivated. This article reviews current outsourcing and collaboration strategies to provide a perspective on how great knowledge sharing could help revise the drug-discovery process. Graeme M Robertson*1 & Lorenz M Mayr2 1University of Perugia, Department of Chemistry & Pharmaceutical Technology, Via del Liceo 1, I-06123 Perugia, Italy 2Novartis Pharma AG, Novartis Institutes of BioMedical Research, CPC/EPP, Fabrikstrasse-16.3.72, CH-4002 Basel, Switzerland *Author for correspondence: E-mail: grobertson@chimfarm.unipg.it For reprint orders, please contact reprints@future-science.com

Basic research Pharmaceutical research Clinical research Skills and competencies available in academia and basic research centres Skills and competencies available at big pharmaceutical and some biotech companies Skills and competencies available in fully integrated companies Academic collaborations CRO collaborations Academic medical centers CRO collaborations Basic research Target validation Screen- to-hit Hit- to-lead Lead-to -candidate Candidate- to-clinical candidate FTIH to PoC Phase II Phase III Launch marketing etc Experimental clinicians Academic industrial collaborations Perspective | Robertson & Mayr Future Med. Chem. (2011) 3(16) 1996 future science group Key Terms Knowledge sharing: Propagation of data and information across research centers to allow the building of knowledge at multiple levels across drug discovery. Collaborative networks: The building of networks of centers of excellence that allow a greater level of achievement by building on in-depth knowledge. Bringing together groups of experts provides a framework for better tackling the challenges of knowledge generation. This article provides an analysis of various approaches to external(and often offshore) out- sourcing activities before reviewing some of the options for more collaborative and innovative interaction frameworks [4]. We will discuss and review the merits and shortfalls of each approach in accessing knowledge and expertise. Innovation & knowledge Drug discovery has traditionally been the domain of pharmaceutical and lately biotechnol- ogy companies, whereas academic groups have, until recently, largely focused on basic research. This division has helped develop an imbalance between drug discovery and the basic biology needed to support it. The current trend for many large pharmaceutical companies to enhance their focus on development has further emphasized this situation and helped decouple the knowl- edge necessary for disease understanding from portfolio decision making [5]. It has been proposed that no pharmaceutical company will be able to ‘profit alone’, but rather, they will need to ‘profit together’, by joining forces with a wide range of organizations, from academic institutions, biotechnology compa- nies, CROs and technology providers, to better arrive at a sustainable approach (Figure 1) [6,7]. Motivations that are used for such collabora- tions range from amalgamating competencies or know-how, to capacity or cost-driven approaches. Outsourcing or collaboration options need, how- ever, to become more knowledge centric if they are to bridge the ‘innovation deficit’ and build interactive interactions between individual cen- ters of excellence that allow access to high-level expertise concomitant with reducing fixed costs that confront but not reduce complexity. Many companies are, therefore, internally focusing on their core competencies while looking to col- laborate with external centers of excellence via not only outsourcing but also via other more col- laborative approaches to access knowledge and thus drive innovation within the drug-discovery process [7]. Outsourcing is often used to convert fixed costs to variable costs. But while appealing for provision of flexibility in resource and cost management outsourcing as a standalone tactic it is not an effective option to increment pro- ductivity [201]. Cost management or short-term responses to shareholder expectations, need to be balanced with a more sustainable business model where these requirements are matched with the sharing of scientific advancement to drive the knowledge generation. Innovative financing or resourcing strategies can expand company port- folios but they will not drive adding value unless they also directly confront the knowledge gap and improve translation of research compounds into viable clinical candidates. Many quote the phrase ‘fail compounds early’ as a mantra for Figure 1. Collaboration and outsourcing strategies, how skills and competencies can be brought together throughout the R&D process. FTIH: First-time-in-humans PoC: Proof-of-concept.

Collaboration versus outsourcing | Perspective www.future-science.com 1997 future science group efficiency in the drug-discovery process, but failure is still failure, and it is clear that sustain- ing the drug-discovery industry on current suc- cess rates is no longer viable. Advancing more programs that fail in clinical trials reduces the overall probability of success for any organiza- tion, no matter how these project are funded or how small the investment. For long-term financial success, the right balance needs to be struck between short-term cost savings, the need to develop, aggregate, and share knowledge, and better integration of basic science into an iterative drug-discovery process. For example, combining knowledge from basic biology developed in academia with medicinal chemistry would develop a better understand- ing of drug/target mode of action and could, thus, promote a more efficient translation of effective compounds into clinical studies and beyond [8]. Many are, therefore, looking to alternative approaches towards reducing costs and improving the science base including out- sourcing and alliances with CROs, risk/reward sharing alliances, precompetitive networks, and academic–industrial collaborations [4]. To support and nurture knowledge gen- eration it is essential that a strong motivation for science is more prominent. Collaborations should not be simple process agreements but rather interactive and mutually beneficial alli- ances, which fuel ‘what if’ questions and help nurture a creative environment. Outsourcing of some specific tasks, such as compound scale-up or GMP manufacturing, are examples where process agreements are valid, but outsourcing of knowledge-generating research activities in a closed one-way agreement will not produce the insights needed to drive innovative drug discovery and will not build sustainability for the industry. Instead, the long-term vision from pharmaceutical companies should be to better encourage, nurture, and engage innovation. Unfortunately, true innovation is often viewed as too risky for many institutional investors who are seeking short-to-medium term returns. Commercial success obviously correlates with a sustainable product pipeline, but the need to stimulate and nurture innovation requires a stronger integration of drug discovery with basic science and a longer-term perspective on where and how best to make investments. The chal- lenges are high but the risk of continuing with the current reductionist approach is even higher. A greater freedom to properly explore novel ideas should be the hallmark of an industry that survives on individual and team creativity [8]. A more entrepreneurial approach and openness to scientific exploration would, thus, provide the innovation that industry requires [9]. Providing freedom to innovate across disciplines and, indeed, collectively across research centers, will help develop a more in-depth but integrated view of what science is needed to better drive drug discovery and respond to the innovation deficit in today’s drug-discovery approaches. Individual disciplines could then diversify at pace with the increasing understanding of chemistry–biol- ogy interfaces [10]. Stimulating innovation in the pharmaceutical industry through contact between research centers be it by, outsourcing of R&D activities, collaboration with academic groups, or alliances with other companies will also help build collaborative information net- works (knowledge banks) with a better balance between efficiency and innovation [10]. Such changes would also have significant consequences for training new scientists in their respective field and help stimulate a new era of scientific discovery also in the academic environment. It should be noted that academia has, tra- ditionally, been much stronger in disease biol- ogy expertise and interrogation of underlying mechanisms, than in medicinal chemistry [11] and the use of chemical information to drive drug discovery [12,13]. With the introduction of molecular biology, new tools (compounds) were required to investigate the modulation and sig- naling of these targets. Thus, developing medici- nal chemistry capabilities within the academic setting was given a new impetus [14]. Coupled with policy changes in the use and ownership of intellectual property (IP) by academic and pub- lic sector institutions (particularly in the USA) this provided a platform for academic groups to engage more fully in downstream phases of drug discovery [15]. However, within academic col- laborations there is also a strong need to promote paradigms that include the formation of inter- disciplinary research teams [16]. This requires a greater collaboration between academic disci- plines to best leverage the advances in under- standing of target modulation [17]. In particu- lar, what is missing is a network of medicinal chemists in academia who can devote themselves to translating promising biology into interest- ing compounds that will help capitalize more fully on the wealth of knowledge on biological systems and the targets therein. The move towards a mixture of (fully) diver- sified (outsourcing) or federated (collaborative)

Perspective | Robertson & Mayr Future Med. Chem. (2011) 3(16) 1998 future science group strategies with some companies exiting from parts of research and development, especially early (basic) research is still evolving [7]. This opens great opportunities for collaborative approaches to establish a greater understanding of the complexity of the respective biological sys- tems. Subsequently, better interrogation of these systems with medicinal chemistry may help to generate a more predictive approach to translat- ing research projects into successful clinical can- didates [7]. However, whatever the approaches adopted they must confront better the barriers to learning and allowing the time to generate sufficient understanding of the system under investigation. This will require longer-term col- laboration to bring the knowledge generated in basic research into a drug-discovery approach driven by science rather than short-term business objectives [18]. Pharmaceutical and biotechnology busi- nesses have always conducted activities globally, so within core R&D activities many companies are using CROs (both in research and in clinical studies) to not only turn fixed-costs in variable- costs, but also to access external skills sets, or technologies. Many factors drive the outsourcing of research and development activities, such as access to specific expertise (for example, animal models or disease expertise in clinical studies), the diversity of scientific approaches offered by these centers, or more tacit aspects, such as access- ing knowledge and understanding. Culture shar- ing and the passing on of tacit knowledge are two fundamental but much less visible elements of R&D alliances [19]. Whether collaboration is via academic–industrial collaborations, precom- petitive collaborations, or public–private part- nerships (PPPs) [20] capturing tacit knowledge [21] will add significant benefits. To benefit more fully from collaborative approaches, companies, and research centers also, need to take the initiative to move towards more adaptive and, hopefully, predictive approaches based on bringing together these centers of excellence to build knowledge and a common understanding of the complex- ity of disease biology. This can be better used in the design of compounds that can alleviate the disease pathophysiology. Indeed greater collabo- ration between the industry, academia, regula- tors, governments, and healthcare providers has been highlighted as one of the main approaches for the pharmaceutical industry to become more innovative and cut its R&D costs [22]. The science and technology of drug discov- ery is advancing rapidly with ever more data available for decision making. However, the overall process of drug discovery in which these activities take place has not changed sufficiently and requires significant adjustments to become a more sustainable concept where preconceptions are more freely challenged and the dynamics of these endeavors embrace more of a disruptive innovation approach [23,24]. If we are to cap- ture and integrate the best science and in-depth knowledge, this will need to be done within a framework where all contributing parties can benefit. Drug discovery is potentially entering just such a period of disruptive innovation. Some have argued for open innovation and this clearly has a role to play, but sustainability will come from a shift to changing the overall ethos and balance between innovation (science) and business (IP) drivers. Innovation & knowledge sharing ‘If we only knew what we already know!’: a much-used phrase, but one that is particularly pertinent in the science of drug discovery [25]. Unfortunately, much, if not most, of the data generated globally are only used to a fraction of their full potential and remain ‘deposited’ in isolated silos rather than used to generate a com- mon understanding. The sharing of knowledge to advance scientific understanding has always been one of the main forces driving science and innovation [26]. Scientific advancement builds on the learning and observations of others and interpretation and application of these observa- tions in new areas or in a novel manner to go beyond the boundaries of previous knowledge. So, with an environment in which the in-depth knowledge of disease biology and, hence, the ability to move towards a more predictive drug- discovery process is inhibited by the over-use of a proprietary IP approach, the pharmaceutical industry is suffering from a restricted sharing of knowledge. This manifests itself in the high level of failures in Phase II when research hypotheses final meet their chosen clinical setting [27,28]. This drive to ‘protect’ commercial interests even from early in research has historically led to a largely one-way process within ‘collabora- tions’ with the data generated being regarded as confidential and proprietary property of the ‘sponsoring’ company. While helping pharma- ceutical (and biotechnology) companies to pro- tect IP around new compounds, this strategy has also helped build a gap between detailed under- standing of compound action against the biology underlying diseases and disease phenotypes [29]. Key Terms Tacit knowledge: Knowledge developed in centers of excellence that derives from the data and processes adopted rather than direct scientific interpretation of the data. Common understanding: The developing of a greater and shared knowledge relating to disease mechanisms and ligand modulation of these processes. With a common understanding, a more predictive environment can be created. Open innovation: Collaboration between centers of excellence in a precompetitive manner in which innovation can be stimulated by a less-restrictive use of intellectual property.

Collaboration versus outsourcing | Perspective www.future-science.com 1999 future science group This has occurred, in part, through lack of data sharing but also to an overly narrow use of data in pushing the compounds forward in single projects and neglecting the retrospective use of the data to build in-depth knowledge for future projects. Reversing this ‘protectionist’ trend with a more widespread collaboration within and between industry and academia, respectively, would help expand the depth and breadth of knowledge and allow the reinvestment of knowl- edge in building better insights into basic dis- ease biology and how to modulate it. A shared infrastructure would help promote fertilization of novel insights between people and research centers with complementary skills. For example, knowledge of the physiological role of the new targets and their relationship, if any, to the etiol- ogy of a disease process can be investigated in detail in academic groups specializing in gener- ating in-depth knowledge. In contrast, drug-dis- covery companies have a breadth of experience in understanding how to progress compounds in parallel versus target biology and relevant chemophysical properties. Obviously, there is a need to align timeframes, but an open collabo- ration between these groups could represent the best of both worlds. It is better to seek to excel in key core areas and strengths and then col- laborate, rather than try to complete the drug- discovery process in each partner individually. For example, academic groups have great flex- ibility and ability to find new knowledge but this can be significantly compromized if they try to engage in the types of high-throughput com- pound screening that industry is set up to do. Better that each center excels at what it is good at and collaborate with other centers of excellence having complementary skill sets. Networks of excellence that have shared responsibility and common objectives and, therefore, consolidate excellence across disciplines are likely to be both more productive and more innovative. This apparently utopic vision is not that dif- ficult to achieve, the information infrastructure exists, but there are significant obstacles to over- come particularly, around accepting that to gain value from the knowledge generated in other centers a more open approach to IP is needed. This knowledge can be difficult to access and develop since the industry–academia interface is still often restricted by closed collaboration models in which the knowledge gained remains confidential [19]. Closed collaboration is the easi- est route towards commercialization of IP but raises significant barriers to learning and the sharing of tacit knowledge. The dominance of this proprietary culture, which tends to value internalization of ideas over creating a common understanding, needs to be overcome in order to find a better balance of more open collaboration that still leads to usable IP. Drug discovery is, by its very nature, an exploration of variable and dif- ficult to predict combinations of data from many different domains whose interpretation requires close collaboration between experts. Within big pharmaceutical companies, this has tradition- ally been achieved by internal vertical integra- tion, while academia and biotechnology have often adopted more network-based approaches. Sharing and collaboration can be produc- tive on many levels particularly if a common goal can be established and the right business anatomy can be achieved [18]. One is precom- petitive sharing of knowledge and resources to make the discovery process more efficient, such as Eli Lilly’s Phenotypic Drug Discovery, PD2 initiative [30,202]. PD2 is currently a 2D model, between the submitting investigator and Eli Lilly, but it has the potential to become a much more innovative model if adopted more widely and is, perhaps, an example of a starting point for vertical integration across multiple organiza- tions. Collaboration can help realize goals that are difficult to achieve and require the bringing together of diverse expertise. This has been a theme of the European Commission Framework programs for research and technology develop- ment, which help to bring together European academic and industrial centers in common goal projects [31]. Within these projects, the promotion of partnering, co-operation, greater complementarity, and collaboration are a pre- requisite. EU Framework projects also help to strengthen the research infrastructure across organizations [203]. This is also a key element of the Innovative Medicines Initiative (IMI) in Europe [32,204] and the Critical Path Initiative (CPI) in the USA [205]. In fact, the European IMI is specifically aimed at strengthening col- laboration between academia, small-to-medium- sized enterprises and pharmaceutical companies across Europe [33]. Integrated consortia can be very effective, but require the full and consistent commitment of all partners and can easily be derailed by inter- nalization of IP by the dominating partner. This mindset of IP internalization, particularly at early stages of drug discovery, is becoming unaf- fordable largely since the knowledge required to