Novartis Vaccines: Innovative Collaborations Leading to Innovative Science

Introduction

Managing research and providing the right structures to allow innovation to flourish is not trivial. Collaboration is one of the essential elements in producing innovation. One of the key levers to producing an innovative group is enabling collaborations. This is even more important in China, where unmet medical needs are actively defined by the government and strong public and private networks dominate. Collaboration provides four levels for innovation:

1. Participants in collaborations are more productive and innovative. Collaborators are more productive and innovative through the act of collaboration.
2. Collaboration exposes ideas to more rigorous examination earlier, stopping unproductive efforts and tempering quality.
3. Collaboration and publishing are highly correlated, and so are grant approvals which fund cross-divisional collaboration.
4. Collaboration enables access to the best scientific resources at a much lower cost than building similar or inferior capabilities in-house. This allows for focus on core competencies and enables rapid access to innovations, of great importance as the majority of innovation is incremental being built on top of what has gone before.

Taking a step back, let us ask why we are interested in innovation. What does it provide? In the case of the pharmaceutical industry, there are millions of unmet medical needs that can be addressed by innovative science. While we have solutions that have dramatically increased human life expectancies and quality of life, anyone who has ever caught a cold or developed cancer will be able to attest to the volume of problems still out there, waiting to be addressed by innovation.

Having established the importance of innovation, I suggest that the stereotypical vision of the innovator as a lonely and isolated scientist toiling away in the lab is not an accurate portrayal of how innovation occurs. Innovation is driven by collaboration, both on a micro-scale in the lab (or more regularly the coffee room near the lab) and on a macro-scale via cross-organizational collaborations spanning academia, industry and governments. Further, looking at historical trends towards convergent innovation (from the discovery of calculus to the identification of HIV), the majority of innovation builds directly on the mass of collectively available information.

The model tying collaboration and innovation is very clear in the context of American and European research efforts. However, its importance is enhanced when you look at the rising situation in China, where the volume of academic talent is supported by a government that proactively defines areas of interest and supports them via funding and oversight.

Collaboration Enhances Productivity and Innovation

Scientific collaboration has been demonstrated to have a direct impact on productivity in both a micro and macro sense. In his studies of laboratory work and productivity, Kevin Dunbar noted that the greatest innovation occurred not in the lab itself, but in weekly lab meetings. Dunbar identified the key cognitive processes that need to go on, “key components of contemporary scientific thinking that are important in generating new models, modifying old models and solving difficult problems”ⁱ.

In addition to these key components there are a specific subset of processes that are important for model building and positively impacted by collaboration, “Analogical Reasoning, Attention to Unexpected findings, Experimental Design, and Distributed Reasoning”ⁱⁱ.

Analogy is actively used by scientists as a method to provide coherence to experimental results and help drive new insight. Collaboration encourages development of new analogies and tests them via other viewpoints. Unexpected results benefit greatly from collaboration as the first instinct of researchers is to discount items that disagree with the current hypothesis (in a study of 223 unexpected findings 196 of these were attributed to methodological issues and only 27 prompted new theoretical models) as experimental error, open discussion of unexpected findings makes it more likely that these findings will be investigated an a new hypotheses developed. This is key since if an unexpected result repeates, a scientist is much more likely to develop new models to explain the results.

Experimental design is clearly influenced by collaboration, but the clearest benefit of collaboration occurs through distributed reasoning. A recent study has shown that as much as 50% of reasoning that takes place in lab meetings is distributed over multiple members of the group. The real impact of distributed reasoning can be summarized by Dunbar:

"What we found is that the major event that occurs during social interactions is generalization. The scientists in the lab often take a finding and attempt to generalize it to other findings, both within and outside the lab. This process is important, as scientists frequently force each other to consider other representations and explanations for the data being considered or theory being proposed. What we see happening is that large changes in representation occur at these moments. One place where distributed reasoning is particularly important is when a series of unexpected findings occur. When a single unexpected finding occurs, little distributed reasoning is present. In this situation, social interactions are usually a straightforward suggestion about method that does not necessitate any major change in the scientists' representation of the issue. But when a series of unexpected findings occur there is a large amount of distributed reasoning. Thus, it is not one scientist shouting "Eureka", but a number of scientists building a new model by adding different components to the model. As noted above, an important feature of this process is that often the members of the lab propose alternate models and explanations for the unexpected findings as well as resulting in the use of more distant analogies. By looking at many different laboratories we also have found that when groups of scientists reason, the diversity of the group is very important (Dunbar 1995). When all the scientists are from the same background it is difficult for them to generate multiple hypotheses, but when the scientists are from different backgrounds many different hypotheses can be generated."ⁱⁱⁱ

We can see at the micro level, that collaboration enhances innovation, but data implies that this is also true on a macro-scale across labs. A study by Zuckerman of Nobel Prize-winning scientists compared to scientists of similar tenure and scientific standing are significantly more collaborative throughout their life-cycle of work, particularly before they win the prize.^iv

This result has been repeated with other cross-sections of scientists and a clear relationship between collaboration and productivity is observed. A key result of cross-lab or institutional collaboration is an increase in diversity, which enables more hypothesis generation.

Collaboration, Publishing, and Grants

1. Collaboration is key for publishing, particularly for quality work. Beyond the direct benefits of collaboration, it enhances publishing, particularly of higher quality work. Those who collaborate more are more productive, and those who collaborate globally are more productive than those who only collaborate locally. In addition, those who collaborate publish more, and with greater impact.^v Many studies have confirmed the findings of Zuckerman with strong correlation between collaboration and publishing.
2. To be able to collaborate with the best, it is essential to publish and enrich the pot.
   In addition to the direct and measurable productivity gains of collaboration for publishing, publishing quality science is a pre-requisite to being able to collaborate with the best scientists. This is of particular importance for corporate scientists, as the best academic and governmental scientists already have access to extensive funding, so to be able to bring value to a collaboration, a corporate scientist needs to demonstrate the quality of their science and the value they bring to a collaboration beyond the simply monetary. The best way to do this is through a record of high-quality publications.
3. Working with the best and producing high-impact publications are essential to receiving grants.
   The ability to work with the best collaborators forms the basis of a virtuous cycle, since collaboration with the best leads to shared publication and makes those involved attractive to the major grant bestowing agencies.
1. Grants produce a special dynamic: The ability to receive grants creates a special dynamic that enables greater productivity and positive externalities.
2. Grant work tends to generate more publications: Grant work tends to produce more publications than non grantrelated work. This is due to a variety of factors, one of which is the rigor applied to the process of receiving a grant already confirms that the area of research is of general scientific interest, thus improving the odds of publication of the related work.
3. Collaboration enables cross-group work that would not occur without the supplementary budget: Funding for collaborations enables cross-functional work that would not be able to occur without the supplementary budget. While many recognize the value of cross-group or organizational collaboration, most groups have internal priorities to absorb all of their internal funding. Dedicated funds for collaboration provide an incentive to do work beyond core focus, and it is this cross-functional work that supports more dramatic innovation, compared to the greater rate of incremental innovation that tends to occur with the focus on internal efforts. While a rational economic view would posit that something worth doing should be valuable whether done with internal or external funds, in practice this is not always the case. Dedicated funds can produce greater benefits and positive externalities via funding of collaboration, which more than offset the transaction cost involved in setting up the collaborations.
4. Funds are the development of key platforms that enable applied innovation: The final element of the value of grants to collaboration and thus innovation, is the ability to establish foundational science and platforms that can then be applied to innovation. It is very difficult for companies with finite resources to fund basic scientific research only tangentially related to their product pipeline. However, grants are frequently focused on basic science questions that are applicable to a variety of applications. External funding that provided the basis of the internet is probably the best public example of the development of a platform that produced tremendous returns to private and public players via applied innovation. Beyond the internet, governments look to fund the basic efforts that will enable new classes of treatments and medicines. This results in a win-win situation – the government is able to address problems it faces, while companies are able to apply the platform technology to develop targeted treatments.

Access to the Best

1. Leading innovators are more willing to work with active collaborators:
   One of the key elements that is often overlooked in the worries about sharing information is that to participate in the broader scientific dialogue, you must give, not just get. As has been shown, the most productive scientists tend to be the most collaborative and the opportunity to work with the best is a function of your own organization's scientific collaborations. Being an active, productive collaborator leads to collaborations with leading scientists.
2. Access to the best enables better innovation by limiting distribution of focus and allowing specific centers of excellence to develop: Access to the best scientists is of increasing importance as the complexity of science has grown exponentially and it is impossible to be the best at all things. Even if it was possible to be the best in all areas of science the investment would be prohibitive and by the time the investment was complete new areas of science would have opened up beyond your purview. By collaborating, scientists and organizations are able to focus on their core competencies gaining an edge in their area of focus, and accessing needed services from collaborators is easy.
3. Switching costs are low as there is not a large upfront investment, so as new technologies arise it is easy to pursue: Another positive aspect of collaboration is that your organization does not need to make the investment themselves in every technology, they can partner with those who have developed the skills and expertise. Since the investment is modest, it is easy to switch or expand collaborations to the areas of science most important to your organization. This is key as new technologies can provide quantum leaps in functionality.
4. Fundamentally working with the best produces more innovation-quality of information, and benefits from lack of overlapping skill sets: The benefit of working with the best is that the best produce more innovation. It also allows for “pressure testing” of ideas with the innovative, ‘out-of-thebox’, which could yield better science and more innovation. The choice is not to innovate together or innovate alone, but to innovate together or not to innovate.

Reverse Vaccinology: A Case Study in Collaboration Enabling Innovation

A case study example of these principles in action is the development of Reverse Vaccinology. Traditionally, vaccines had been developed according to the empirical process of isolate, inactivate, and inject. While this was an effective approach for many early vaccines, this method is not applicable to all vaccines, in particular complex bacterial pathogens. With whole genome sequencing came a new opportunity for the development of vaccines.

However, in the 1990s, the ability to sequence whole pathogen genomes was limited and not available to many vaccine companies. Scientists from Siena, Italy, reached out to Craig Venter and TIGR, for the potential of sequencing the entire genome of troublesome pathogens. They started by targeting Meningococcus B, a lethal pathogen not addressed by a standard polysaccharide vaccine due to the polysaccharide in the Men B coat. Through the collaboration with TIGR, the entire genome of the Men B pathogen was sequenced. Then, the various genes were expressed and tested as potential vaccine antigens. Leading candidate antigens were identified, refined, and became the basis of a late-stage vaccine candidate.

Figure - 1

Having been applied to many pathogens, this approach has proven very fruitful. Reverse vaccinology has entered common academic use after its publication. This is a case that clearly demonstrates how collaboration can lead to a breakthrough that was otherwise impossible and shows that and the fruits of that collaboration lead to other beneficial partnerships.

The Situation in China

The benefits of collaboration are clear for Europe and the US, but the question is: how applicable is this model to R&D in China?

The answer is very applicable, because of the shared factors that make it effective in the US and Europe, and because of specific features unique to China.

1. Active role of government in determining problems: The government in China plays a very active role in determining key unmet medical needs, through its surveillance network with the China CDC, the SFDA, and Chinese Grant making authorities. These groups work together to identify issues, fund solutions, and enable approval.
   1. Problem-definition enables more targeted and productive innovation - necessity is the mother of invention: By defining the key problems and unmet medical needs, the Chinese government creates a set of limits that enables more innovation. A totally undefined scope is less likely to determine innovative solutions than a problem with at least one partial limit on scope.
   2. Government actively supports collaboration via grants, and a willingness to invest in the human capital benefits collaboration: China has strong collaborative roots, and the grant-making culture encourages collaboration. In addition, China is aware of the human capital benefits of collaboration. When senior scientists devote time and resources to collaborations with more junior scientists, both reap benefits, greater productivity, and enhance total Scientific Human Capital as their collaborators are developed by their expertise.

Conclusion

Collaboration is key for enabling innovation, and organizations that are open to external collaboration for their core products will take advantage of this. Openness to grants is also key as they enable collaboration and push towards better science.

ⁱ Dunbar “How Scientists Build Models InVivo Science As A Window On The Scientific Mind”
ⁱⁱ Dunbar “How Scientists Build Models InVivo Science As A Window On The Scientific Mind”
ⁱⁱⁱ Dunbar “How Scientists Build Models InVivo Science As A Window On The Scientific Mind”
^iv Am Sociol Rev. 1967 Jun;32(3):391-403. Nobel laureates in science: patterns of productivity, collaboration, and authorship. Zuckerman H.
^v Pao, Miranda Lee “Global and Local Collaborators – Publication Drives Funding” 1992 Information Processing and Management