Pooling multiple pharmaceutical companies’ assets across a given therapeutic area will allow industry to make rational R&D investment decisions, expedite resource allocation and reduce investment in also-rans, eventually boosting productivity and improving a worsening reimbursement climate. Oncology may be a smart place to begin.

The pharmaceutical industry is no stranger to innovation—its very survival depends on it. In one area, though, the industry has been surprisingly slow to change: its business models for R&D. Despite rapid change affecting many aspects of the markets, customers and regulations that define the operating environment for pharmaceutical firms, the core innovation and product development model has remained largely the same for the last 25 years. Today’s R&D technologies are much more sophisticated but the fundamental resource allocation and risk management models in place at many large pharmaceutical companies remain rooted in the past.

It’s time to evolve. Pharma companies need a new mechanism to make rational R&D investment decisions across the entire industry portfolio—rather than on an isolated basis. A model based on partnering is one way to increase the efficiency of research spending. By effectively “pooling” research assets within a disease area, the industry would benefit threefold. One, industry players would make resource allocation decisions much earlier than before and fund only those projects with the highest likelihood of showing true clinical differentiation. Two, this approach will reduce the number of duplicative or “me-too” products in the market struggling for reimbursement. Finally, the new model will distribute risk and returns across partners to increase the predictability of pipelines and long-term revenues.

It might seem radical, but for some time now, market forces have nudged the industry to collaborate on R&D. Historically, pharma companies have pursued drug discovery and development on their own, bearing all the risk but also reaping all the rewards. That model relied on low-hanging fruit for drug discovery and development, plenty of resources to invest in all the projects in the pipelines, and payers willing to reimburse most approved compounds regardless of their clinical differentiation from existing therapies.

Much of that no longer holds true. Executives are acutely aware that more than $48 billion of drug revenues will come off patent cumulatively in the next four years. According to Bain analysis, $41 billion of the $48 billion are revenues from blockbusters that typically produce operating profit margins of 60 to 75%. With blockbusters facing competition from cheaper substitutes, nearly $30 billion of yearly cash-flow will drain out of the pool. That $30 billion is equivalent to 50% of the annual R&D spending by pharmaceutical companies. It puts the challenge in stark terms: Senior pharma executives will be compelled to do more with less capital, with very little industry-specific experience to draw upon. This, in part, explains the recent rash of merger announcements—Pfizer Inc.-Wyeth and Roche’s hostile offer for Genentech Inc.—while other large pharma companies, including Sanofi-Aventis and Merck & Co. Inc., have indicated that they too are open to large scale acquisitions.

But are mergers and acquisitions the cure for what ails the industry—the need to make the R&D business model more efficient? Not really. Mergers and acquisitions such as the recent Pfizer-Wyeth deal take place for many reasons—minimizing costs, diversifying the portfolio—but they do not guarantee reducing risk or improving the productivity of R&D. In fact, most mergers have yielded very little gain on the R&D side. Managements tend to retain most of the pipeline, physical assets and R&D talent of the combined companies—which makes decision-making even more difficult. Often M&As work only in the short-term and do not help in substantially increasing R&D efficiency or redistributing risk, for the two merged companies, or the industry as a whole.

Consider the evidence. Despite the wave of major mergers that began in the mid ‘90s and created the current Top 10 pharmas, industry R&D productivity has suffered dramatically. Pfizer has seen a continuous fall in R&D productivity since it subsumed Warner Lambert in 2000 and Pharmacia in 2003 despite being the industry’s biggest spender on R&D. With this track record in mind, not everyone is hunting. Despite low valuations and the temptation to look for deals in the current economic outlook, pharma companies like GlaxoSmithKline PLC and AstraZeneca PLC have announced that they are not inclined to do large scale mergers given the uncertain value for their shareholders.

Instead of acquiring resources, a more sustainable solution for Big Pharma lies in pooling resources. In the near term, more collaborative business models for R&D can improve the ROI of the drug development process (which we currently estimate at approximately 4% versus 9% during the 1995–2000 period), lower costs, speed time to market, spread risk, and improve market potential.

Most importantly, collaboration can help reduce wasteful overlap. In a number of primary care diseases, including gastroesophageal reflux disease, hypertension and hypercholesterolemia, many pharma companies have focused intensively on the same opportunities. And for compelling reasons: there were a large number of patients, the diseases lent themselves to the development of new drugs, and there was scope to use aggressive sales and marketing to build market share. However, as overlap increased, the constraints of the existing model became apparent. High investments continued despite decreasing probabilities of success, duplicate investments increased with companies investing in very similar products, and an increasing number of players began crowding the same narrow space. As a consequence, in these disease areas, innovation productivity declined; there was very little differentiation between products; and commercial costs increased substantially. It also represented a lost opportunity: Payers are more likely to pay a premium for products that are clearly clinically differentiated.

Oncology: Ripe for Asset Pooling

Oncology is the latest disease area where these trends are increasingly apparent. While the number of patients is smaller in oncology than most primary care segments, many pharma companies have been attracted to the same opportunities due to several reasons: smaller commercial investment requirements, significantly higher prices for oncology drugs, payers who are willing to pay these prices, and a rapidly growing set of potential new drugs. The number of preclinical and clinical oncology compounds in pipelines is two to three times that of the next largest disease area. Pfizer, Eli Lilly & Co., Merck, GSK, Takeda Pharmaceutical Co. Ltd., and Eisai TK have made big bets on oncology in quick succession while more traditional competitors like Genentech, Bristol-Myers Squibb Co., and AZ continue to increase their investment. In the last five years, Pfizer has increased its oncology R&D projects by 400%, created a specialist oncology business unit, and now has more than 20 oncology compounds under various stages of development.

Only a limited number of winners will emerge in oncology—there is not room for everybody. The current development pipeline is jammed with overlapping products across oncology pathways. More specifically, if the number of pipeline products with indications by tumor type are considered, theoretically, 40 to 50 compounds could reach the market over the next 10 years, given the current probability of success. But the existing model offers no way to rationalize these duplicate investments until the products reach the market: many candidates are developed in strategic isolation and there is no mechanism to make informed decisions about comparative efficacy.

By that point, huge investments will have been made, although only a few products will have differentiated clinical efficacy and will emerge as market leaders. Only these compounds are likely to command premium prices and meaningful market share; most will struggle to recoup a return on investment as reimbursement pressures inevitably spread from primary care into specialty care. The endgame for oncology is heading to the same point that primary care finds itself in now: low returns on R&D, a crowded competitive landscape, significant price competition and higher commercial costs.

Changing the stochastic nature of the drug discovery and development process is a very tough problem to solve—if it ever can be solved at all. Nor is technology the panacea. The industry had high hopes first from throughput screening, then combinatorial chemistry, then genomics. None of these markedly increased efficiency or reduced costs in R&D. On the other hand, the new approach of pooling resources to create real value could impact both those issues—especially in crowded areas like oncology.

The Mechanics of Pooling

In the collaboration R&D model, a “seller” would license an entire pipeline of products within a disease area or class, say oncology, to an “acquirer”. In return, the seller would get a share of the combined pipeline output irrespective of whether the seller’s assets reach the market or not. Logically, for this model to work, the acquirer will need to have strong capabilities in R&D and commercialization in the given disease area; the seller may augment these capabilities with some unique capability of their own such as a specific technology platform or insights on developing a specific biological target.

Although new to the pharma industry, such collaborative models have worked well in bringing competitors together for the common good in other industries. In consumer electronics for example, firms often come together in successful consortiums to develop new technologies. Video storage formats like Blu-ray, HD and DVD achieved commercial success only when industry players ultimately agreed to work together. Similarly, the NEC, Sony and Toshiba collaboration helped develop the 45 nanometer micro chip at a lower combined cost for the industry, while allowing each company to retain its own competitive edge. In the 1920s, the manufacture of airplanes would have stalled if Congress had not created a “patent pool” to nudge companies to share technology—and allow a nascent aircraft industry to take off.

What sets this model apart for the pharmaceuticals industry is that the value created is not driven purely by the partnered products—but rather by the authority over resource allocation to pursue the most-likely-to-succeed products from a larger candidate pool. By putting the resource allocation authority under one umbrella, this model helps to ensure that decisions are made rationally and resources are allocated efficiently.

If done right, such collaborative development will help the pharma industry in several ways. First and foremost, it will help eliminate duplicate investment by consolidating pipelines of different companies into a smaller number of broader product portfolios. To be sure, there will still be the need to develop multiple products to service niches—in areas like breast cancer, depression and schizophrenia different products are required for different genetic make-ups. But, over the long term, as personalized medicine leads to more tailored products, pooling will allow pharma companies to make better investment decisions earlier.

The new model will also make allocating resources more efficient: R&D spend can be limited to those products which show the greatest potential for clinical differentiation. In the current silo-ed approach, resources are wasted because comparisons of clinical data can’t happen across company pipelines. Products with less compelling clinical data continue to receive development dollars based on the sponsor’s inability to determine its comparative effectiveness against other industry pipeline products. Other times, products continue to be funded because the sponsor lacks other viable products in the pipeline and feels compelled to push mediocre products forward. In the new model, investment in the weakest products can be terminated much earlier in the development process, and resources can be focused on the highest potential subset of the larger combined portfolio. Finally, the pooling model will also free up resources for industry players to invest in a broader range of therapeutic areas, as R&D advances enable new breakthroughs.

Shaping the New Model

These new partnering arrangements are likely to take several forms, but they will share some common characteristics. All the sellers who participate in the collaboration arrangement will share in the rewards when products from the combined pipeline reach the market. It will be important to ensure that the value-sharing terms eliminate any inherent bias on which compounds are pushed forward from the combined pipeline. Finally, in the new model, the extent of the rewards will be commensurate with the initial formula agreed upon while forging the collaborative arrangement. For collaboration to work, it will be critical to identify the right deal structure.

One reason why Big Pharma is proceeding cautiously is because they know that companies that choose to work together in this fashion will need to define the formula for revenue sharing and investment contribution up-front as well as set clear guidelines for decision-making around resource allocation. All those are tough tasks, but there are several ways these deals might be structured.

In one scenario, two or more companies could pool their assets into one focused pipeline and based on the initial valuation, share the benefits later. Imagine if two industry leaders were to collaborate. In this scenario, the first firm would contribute its early-stage oncology molecules and the two partners would calculate a market value which, instead of a licensing agreement, would be spread over the life of the entire portfolio being developed by the second firm. The first firm would get a percentage of whatever comes out of the pipeline whether it's a molecule contributed by the firm or not. The advantage to the second firm is that it reduces net risk and increases its ability to make portfolio allocations. For both companies, the total amount of spending on R&D goes down and the rewards are shared.

Critical for the success of such collaborative arrangements is the ability to share decision-making, and to clarify upfront which partner has the authority to select the most promising compounds as well as to allocate resources between potential winners. Roche achieved this outcome indirectly with its $2.1 billion investment in Genentech, in 1990. Leveraging the biotech’s expertise in early stage research and development, Roche helped pick the most promising drug candidates for development and took on the responsibility of commercializing the products outside the US. While taking a majority stake in Genentech was not precisely a pooling of assets, it did allow the two companies to focus on what each did best. The arrangement helped both companies avoid duplication and create value—but mainly because Roche allowed Genentech to operate independently. In general though, the minority or majority stake approach is less viable than pooling resources simply because of size: even minority stakes between the top 10 pharma companies would be very expensive and not solve the governance issues required for joint decision making. In the case of smaller strategic alliances like GSK-Theravance Inc., there are visible benefits of pooling, where GSK is able to make pipeline decisions across a broader range of drug discovery programs.

A second scenario could feature external funding. Goldman Sachs’ recent proposal to provide funding for targeted co-development of specific medicines represents another step in the direction of promoting collaboration between competitors. The plan would create a “research pool,” with pharma companies invited to place a range of experimental drugs in a single therapeutic area in early-stage Phase 1 and 2 trials in the pool, and specialists from different companies working together to develop specific medicines. Designed to help pharma companies “trapped in their pipelines,” the new model would create more flexible, transparent and cheaper drug development by allowing competing pharma companies to pool their resources.

Finally, a start-up structure, where pooled investment resources are tied to clinical development programs, might work particularly well in emerging areas of research. Each partner in the licensing arrangement has the right of first refusal, followed by a bidding process for the majority ownership stake. Collaboration would have helped prevent the kind of duplicative (and ultimately non-productive) spending that came about with the development of dual PPAR agonists, for example. Several companies including Merck, BMS, AZ and Takeda invested separately late into Phase III with very similar compounds—and all failed. At the last minute, in Phase III, BMS struck a deal with Merck—but the partnership was mostly related to sharing commercial resources. Under a collaborative model, the industry could have pooled resources and worked together to develop dual PPAR agonists. To be sure, dual PPARs would still not have worked, but the total investment to figure that out would have been dramatically less.

Litmus Tests for Success

The new model could create more winners and fewer losers in a pharmaceutical industry currently facing an uncertain outlook. If implemented thoughtfully, the model is capable of reducing total R&D investment in the near term without compromising the development of drugs in the long term. Still, some may worry that this model will result in less innovation if pharma companies don’t conduct R&D independently of each other. Instead, a more likely outcome is that this model will help create a healthcare system that develops innovative medicines at a more affordable cost. A partnered R&D model is much more likely to yield a more efficient allocation of research dollars, especially in areas which are crowded with duplication today. It will also free up money for pharma companies to invest in making novel, disease-altering discoveries.

That is not to say developing collaboration in R&D is without its challenges. First, there are several thorny external issues that potential collaborators will have to resolve before their partnership gets off the ground: How to do the valuations? Where to find the right funding? How to assign control over resource allocation? Then there are internal obstacles to cross-company collaboration which are equally significant. Big Pharma leaders are the first to admit that management teams find it hard to collaborate with perceived competitors, culturally and viscerally.

But here there is hope. While it is difficult for R&D directors and other senior executives to see pipeline-partnering as a viable approach, it is not impossible. In fact, in the past, Big Pharma has often invested in smaller pharma companies with a limited pipeline of compounds. The takeovers have worked well with Big Pharma weighing those compounds into its R&D investment allocation. Now, Big Pharma has to enter into collaboration on a much larger scale—with other large industry players.

The biggest impetus to refine and adopt this new model will come from market forces. The tightening capital crunch coupled with escalating development costs are forcing even the most independent-minded R&D director to admit that the industry needs to move to a superior model for allocating resources for drug development—and that it is unrealistic to wait for a technology breakthrough to reinvigorate the company’s R&D. The payers’ rapidly declining tolerance for me-too products is also creating pressures that favor collaboration.

A more tacit concern Big Pharma has with the new approach is that it will push all of them to the same mean performance. The pharma companies that have begun exploring the new approach, however, have so far focused their efforts on areas where they have a competitive advantage or strong differentiation. Indeed, this is one of strongest rationales in favor of the new model: it enables the companies with the strongest capabilities in a particular class of compounds to focus on that area of discovery and development.

To get to that winning state of mind requires a willingness to give up a measure of control and be willing to work with a perceived competitor—and that is hard. The early collaboration efforts indicate that successful deals happen between companies that are willing to engage with their partners to manage clinical development and commercialization processes. And more importantly, collaboration works when there is no compulsion to duplicate resources to shadow what the partner is already doing. To a large degree, the extent these deals will succeed or fail will depend on how much each pharma company is willing to eliminate its internal duplicative products once it has entered into a collaborative arrangement.

As more pharma companies begin to see the value in coordinating products at different stages of development in their pipelines, more R&D partnerships are likely to succeed. Already, the dots are getting closer. In the last few years, many executives have come to realize that the current model needs a substantial overhaul, that capital is too scarce a resource to be wasted on inefficient pipeline development, and that partnering with the competition can generate better long-term gains. Some are beginning to act. Earlier this month in a speech at Harvard Medical School, GSK CEO Andrew Witty publicly recommended that drug companies contribute patents to a common pool to accelerate the development of drugs for neglected tropical diseases. If necessity is the mother of invention, clearly, it’s time pharma companies began inventing a new R&D model: with resources running scarce, business innovation must come to the lab.

Preston Henske and Tim van Biesen, PhD, are partners at Bain & Company. They are based in New York and are leaders in Bain's healthcare practice.

Moving R&D toward a collaborative innovation model

1. The first step is a rigorous, objective review of the pipeline to decide which projects are strategic and which are likely to result in products with little or no differentiation from competing products. It’s critical to bring the payer perspective to the table as part of assessing the true market value of the product.
2. Second, companies can assess alternative scenarios of collaboration for the area of highest potential. That will require building new skills in the company’s business development team to identify collaboration opportunities and structure equitable terms for all partners.
3. Third, pharma companies will need to adopt new governance models for collaboration to minimize the redundant investment and high management overheads that currently burden joint-ventures and partnerships.
4. Fourth, Big Pharma needs to significantly improve resource allocation models and risk management models—areas that are currently lacking in discipline. An evaluation of the true ROIC for each program is critical to make the appropriate trade-off decisions.
5. Finally and perhaps most important, pharma companies will need to extend a hand of partnership in a highly competitive industry.

Spiraling Costs of Development

In most industries, the costs of product design and development are rarely an impediment to getting market access--the battles for supremacy are typically won and lost in the market place. However, in pharmaceuticals, even the most promising products face a massive resource challenge to reach the market. Due to escalating costs and falling success rates, fully loaded, risk-adjusted costs of product development and launch have increased to $2.2 billion per product, a more than 25% increase since 2002 and twice as high as the late 1990's. The IRR on these investments has dropped to nearly 4%. With costs unlikely to fall in the foreseeable future, it is more important than ever for scarce R&D resources to be allocated to the most promising products. The silo-ed nature of industry pipelines today stands in the way of efficient investment decisions. These duplicate investments must be dramatically reduced to increase the overall R&D productivity of the industry.