The answer to stagnating R&D can be found in the creativity of the movie industry.
By Liam Bernal

The pharmaceutical industry should be a tangible demonstration of the value of the scientific enterprise. Enormous benefits to society should, and generally do, flow from a healthy pharma sector. The transformation of AIDS into a treatable chronic disease is an example of such a benefit.

However, despite inexorable progress in biologic research, there has been a perplexingly dismal return in the number of drugs that have been successfully developed and marketed in the last few years: The number of new drugs that the US Food & Drug Administration has approved has fallen by half since 1996, with only 20 approved in 2005. That was the second lowest number of approvals in the history of the agency. Failures in the last stage of clinical development, Phase III, are...

Pharma's productivity woes are a consequence of two distinct factors. One, which is difficult to change, is the dearth of "low-hanging fruit" now that easy pharmacologic targets have been exploited. The other, more malleable factor is the current managerial structures in pharma that not only do not foster creativity and innovation but also tend to extinguish it.

The introduction over the last decade of drugs with novel mechanisms of action for lung disease illustrates the low-hanging fruit hypothesis. Though most R&D work is focused on diseases that may provide the biggest potential revenue, the preponderance of new drug approvals have not been for these indications. Asthma is common and represents a huge commercial opportunity, yet only two novel drugs have been introduced for its treatment in the last decade: Singulair, a leukotriene receptor antagonist, and Xolair, an IgE antagonist. A great need exists for better therapies for chronic obstructive pulmonary disease (COPD), yet despite the large potential market, no novel therapies have been introduced in recent years. Moreover, few novel targets have a compelling scientific rationale.

In contrast, primary pulmonary hypertension (PPH), which is rare, was relatively neglected until the mid-1990s. Since then four drugs, each with different mechanisms of action, have been approved for its treatment. The pattern of drug approvals for asthma, COPD, and PPH implies that the main determinant of success is more likely to be the tractability of the specific pharmacologic target rather than the magnitude of the commercial opportunity: It has been easier to develop drugs for the PPH targets. A similar story is seen in other therapeutic areas. For example, of more than 100 drugs tested in clinical trials for stroke following extensive and encouraging evaluations in animal models, not one has been shown to be effective in patients.

It is hard to escape the conclusion that the targets of the future will be more difficult to modulate pharmacologically, and that significantly enhanced creativity and innovation will be required to develop compounds successfully.

The second critical and fundamental cause of pharma's productivity problem, which fortunately is potentially remediable, is what former R&D director for Burroughs Welcome, Glaxo, and Warner Lambert, Pedro Cuatrecasas, has referred to as the "pervasive mismanagement" of the R&D process. Cuatrecasas noted in a recent article in the Journal of Clinical Investigation 1 that the rot started in the early 1970s when managers with business school or legal backgrounds, but no significant foundation in science or medicine, began to invade the upper echelons of pharma and introduce structures and practices such as "management by objectives" from industries lacking any significant R&D enterprises. This invasion was motivated by a desire to increase the efficiency of R&D and to prioritize maximizing the return on investment.

An even more stifling trend has been the recent importation of the "six sigma" business improvement methodology into aspects of pharma R&D. Six sigma was designed to improve manufacturing processes, but has been well documented to quench innovation. 2 The intellectual bankruptcy typical of many current pharma leaders is well illustrated by the typical pharma response to faltering productivity and the resultant fall in earnings. Take, for example, Pfizer's acquisitions since 2000 of Warner-Lambert and Pharmacia. Rather than investigating and addressing the fundamental etiologies of the problem and contrary to the readily available data in the business literature, the leadership plunges into the short-term fix and ego-satisfying drama of a merger, which is almost guaranteed to stifle innovation even further.


To address mismanagement in pharma, what might be learned from the lessons of other industries? Hollywood is dependent on creativity and innovation, so its evolving business model can provide useful insights.

Let's consider analogies between pharma and Hollywood: A screenplay could be considered analogous to a drug.

This is not as crazy as it would seem. The Hollywood business model is more evolved than the model currently followed across pharma. Until the 1950s Hollywood studios employed everyone, from actors, writers, and directors to producers and marketers. Then the "talent" - actors, writers, and directors - realized that their expertise was undervalued, and their personal interests were much better served if they functioned autonomously. Hollywood studios are now mainly responsible only for financing and marketing movies. The creativity resides outside the studios, with independent producers, such as the Weinstein brothers at Miramax. They have created and followed a simple formula that has led to multiple movies, made on relatively modest budgets, that have been both financially and critically successful: Plot must deal with a core aspect of the human condition, central characters must have redeeming qualities to make them sympathetic, and the budget must be rigidly managed.

Let's consider analogies between pharma and Hollywood. A screenplay could be considered analogous to a drug. A compelling screenplay generates a quest to get a director with the best credentials to develop it. The chosen director then personally picks his key colleagues, generating a creative team with relevant complementary expertise and a compatible working style.

Contrast this with pharma: Constructing a team is much more of an ad hoc phenomenon, with little attention given to selecting a team leader and team members. Given the complexity of drug development, individuals with the appropriate constellation of scientific, technical, and managerial skills are rare. They should be afforded star treatment.

But Big Pharma continues to follow the old studio model, though there are signs that this may be changing. A similar and necessary evolution to what Hollywood underwent in the 1950s may be beginning, with increasingly more drugs being discovered outside Big Pharma, presumably because the R&D process elsewhere is more conducive to creativity. Biotechs or small pharma settings tend to be flexible and nonhierarchical, with a tolerance for mistakes and constructive dissent - all characteristics of environments that nurture creativity and innovation. Consequently, the trend towards more drugs being discovered outside Big Pharma is likely to accelerate.

There is a precedent for pharma emulating Hollywood: Pharma's main preoccupation, the creation of blockbusters, was directly copied from Hollywood. The blockbuster model is really defined by broad and aggressive marketing, though the term is less accurately, if more commonly, used to define revenue thresholds. Hollywood's blockbuster model was created in 1951, when the term was first applied to the movie Quo Vadis because of its then-huge budget of $7 million and the unprecedented zeal of its promotion.

Glaxo created the first pharma blockbuster in the 1980's with the ulcer medication, Zantac. The drug was a modestly effective but very safe compound that ultimately attained sales of more than $2 billion a year as a result of extremely aggressive marketing. Though the blockbuster model continues to serve the movie industry well and has undoubtedly worked well for pharma, the recent Vioxx debacle and the demise of Pfizer's torcetrapib (an inhibitor of cholesterylester transfer protein resulting in higher HDL cholesterol levels and reduced LDL cholesterol levels) may indicate its viability and usefulness is now in question. Indeed, the blockbuster model has become a liability.

Though it is unlikely that the best pharmaceutical "talent" will become free agents, the creative class will migrate to smaller enterprises where there is more scope for innovation. To retain such people, pharma might look at the studio with arguably the most highly trained employees, the most collaborative culture, and the best record for consistently producing successful movies over the last decade: the animation studio, Pixar (which Disney recently took over).

Friction between marketers and creative talent is common in Hollywood and pharma.

Pixar nurtures interactions between employees working on different projects with the aim of amplifying the capabilities of teams by fostering a naturally collaborative environment. The Pixar University, which allows employees to spend four hours a week on company time studying artistic subjects unrelated to their actual job roles, is a major factor in educating employees about their colleagues and providing a framework for collaboration on subsequent projects. This organizational culture contrasts vividly with the psychic and structural silos that characterize Big Pharma; there, command-and-control leadership is generally favored.


Friction between marketers and creative talent is common in both Hollywood and pharma. The difficulty of predicting the commercial success of a movie is well captured by screenwriter William Goldman: "No one knows anything." While this phrase doesn't (yet) have currency in pharma, that shouldn't be interpreted as meaning that pharma commercial predictions have good record. Indeed, from Prilosec to Lipitor to Rituxan the record of commercial predictions for early-stage compounds is abysmal. Frankly, the less that marketing has to do with early-stage decision making, the better. Genentech has, in the last few years, easily transcended any other company in either pharma or biotech. This has been attributed to having an accomplished scientist with substantial industry experience as CEO and prioritizing early development decisions based on the robustness of the science and an assessment of medical need.

An openness to new script lines, alternative scenes, and novel cinematography techniques is essential in movie development. However, even though drug development is on average a 10-year development process, as opposed to three for a movie, it often seems to proceed blindly down a predetermined path. The constant accrual of new data requires an ongoing assessment of the benefit/risk profile of a compound to be made and for predictions of future blockbuster status to be adjusted downwards. These decisions are difficult to calibrate and can have profound sequelae.

Inflexibility "a culmination of all the other shortcomings in pharmaceutical development" is perhaps the major hazard in clinical development. One would think that contemporary scientists would be prudently open minded, especially in settings where the stakes are as high as they are in the pharmaceutical industry. But, in part because of the constricting structure and practices in the industry, integration of novel data into subsequent decisions is poor. This is a fundamental cause of the Vioxx fiasco that tarnished the reputation of the entire pharmaceutical industry.

What do we do when we become aware of novel information that conflicts with our established beliefs? We do what the late economist J.K. Galbraith said: "Faced with the choice between changing one's mind and proving there is no need to do so, almost everyone gets busy on the proof."

Though it is unlikely that the best pharmaceutical "talent" will become free agents, the creative class will migrate to smaller enterprises where there is more scope for innovation.

The case of Vioxx provides a unique window into the complex decisions made when developing a new drug. The company's deliberations are open to scrutiny because of the huge amount of data uncovered by an investigation commissioned by the company and conducted by an independent law firm (available at www.merck.com/newsroom/vioxx/martin_report.html).

When Merck acquired clinical data from the VIGOR trial that questioned the cardiovascular safety of Vioxx, the company conformed to Galbraith's cynical perspective. They proposed and ultimately accepted an explanation for the findings that, instead of acknowledging Vioxx as potentially problematic, implied that the competitor drug, naproxen, possessed previously unknown cardiovascular protective properties. Though this explanation was a plausible hypothesis, Merck acted as if it were proven.

The FDA was much more circumspect. And independent experts were not convinced that the outcome could definitively be attributed to a protective effect of naproxen. In fact, experts that the company consulted and who had access to all the available data, said that there was a 15% chance that Vioxx itself had cardiovascular safety problems. Instead of acting in accordance with that probability, Merck acted with conviction that the available data was overwhelmingly in favor of Vioxx being neutral with respect to cardiovascular safety.

Why did Merck disregard or at least minimize this perspective? The company clearly felt that the risk of Vioxx having intrinsic cardiovascular safety problems was substantially less than 15%, because they conducted long-term placebo-controlled trials and confidently stated that these would establish the safety of the drug. One of these studies, the APPROVE trial, subsequently confirmed the safety risk, which then led to the abrupt withdrawal of the drug. Given the profound consequences, why did Merck so confidently underestimate the risk following the VIGOR trial? To address this, it is useful to consider the management style and personality type that dominate Big Pharma.


Berkeley psychologist Philip Tetlock studied 30,000 quantitative political predictions made by 300 experts and categorized psychologically into two classes in his book, Expert Political Judgment: How Good is it? How Can we Know? (Princeton University Press 2005). One class, considered decisive, incorporated what they learn into their established core view, whereas the other, flexible group modulated their thinking as they learned. He concluded that the latter make significantly better predictions about future outcomes because they are able to recalibrate their judgments in accordance with new information. The former concentrate on evidence that confirms their established perspectives.

At Merck, management characterized by a strength and decisiveness made the decisions regarding the significance of the outcome of the VIGOR trial. More flexible leadership would have placed greater emphasis on the FDA's and outside experts' more measured interpretation of the data. Such leadership would have recognized that it was impossible to conclusively exonerate Vioxx based on the available data. Moreover, it might have recommended changes in strategy such as abandoning direct-to-consumer advertising and leveraging the unique proven gastrointestinal safety advantage to high-risk populations. Continued adherence to the blockbuster plan was a strategy blind to the possibility of a real problem, which ongoing placebo-controlled studies would reveal. If Merck, like Hollywood, had used flexible decision-making when they first obtained unequivocal evidence of cardiovascular safety issues in the VIGOR clinical trial and abandoned their blockbuster strategy, the company would now be much better off fiscally, and have a better reputation.

This kind of expertise, wisdom, leadership, and good judgment are not acquired overnight. A study of the 31 major Hollywood studio heads between 1936 and 1965, including such legends as Louis B. Mayer, Darryl Zanuck, and Jack Warner, found that it took an average of 15 years before leaders achieved their highest level of performance 3 . Just as Hollywood needed such leadership then, pharma desperately needs wise leadership now. The storms that it is presently encountering are only likely to intensify.

"Faced with the choice between changing one's mind and proving there is no need to do so, almost everyone gets busy on the Proof." - J.K. Galbraith

In order to cope, companies will need to reassess their approach to developing, attracting, and retaining talent. At present, pharma's approach is stuck in a rut of benchmarking itself only against other pharma entities. An opportunity exists for emulating successful practices derived from other companies such as Pixar and deliberately creating settings conducive to collaboration in order to attract and retain talent and build competitive advantages. For example, Genentech's single R&D campus, which allows many researchers to work on pet projects, is one such setting, but no company has really begun to consider development as a place for innovation and creativity. The reason may be regulatory issues, but it is also because R&D has been treated like a commodity until recently. Other industries that are heavily dependent on the intellectual horsepower might also provide useful models for pharma to consider borrowing; they include the software industry, investing banking, and management consulting.

Boards of directors need to have a much more sophisticated approach for selecting leaders. Selection is a probabilistic endeavor. It is more important than ever that a leader's ability to make decisions be based on an understanding of probability with a capacity to recalibrate one's perspective in the light of novel information- so-called Bayesian thinking.

The best leadership bets will have a scientific pedigree, at least a decade of industry experience, and be openly demonstrable to new ideas and data. The companies that are able to identify and develop leaders with superior decision-making ability will have a compelling competitive advantage in the future. The simplistic days of resolute leaders sticking to key decisions through thick and thin must be avoided if pharma is to survive and regain its credibility. Analysis, continuous learning, and refinement will give pharma leadership the greatest likelihood of success in effectively handling the surprises that drug development will inevitably encounter.

*The author of this piece is a current pharmaceutical R&D executive who has chosen to write under a pseudonym.

1. P. Cuatrecasas, "Drug discovery in jeopardy," J Clin Invest, 116:2837-42, 2006.
2. M.J. Brenner, M.L. Tushman, "Exploration, exploitation, and process management: the productivity dilemma revisited," Acad Management Rev, 28:238-56, 2003.
3. D. Miller, J. Shamsie, "Learning across the life cycle: experimentation and performance among the Hollywood studio heads," Strategic Management J, 22:725-45, 2001.

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?