AFG Venture Group Dispatches

Corporate advisory and consultancy in Australia, South East Asia and India.

What Is The Future For The Big Pharma Model?

Phil Kearney
Director of Licensing & External Research, Merck, Sharp & Dohme

The 2010 Pharmaceutical R&D Factbook, compiled by CMR International, painted a gloomy picture of the global pharmaceuticals sector. Most notable was the finding that new drugs launched within the last five years accounted for less than 7 percent of industry sales in 2009, down from 8 percent in 2008.

“The latest data shows that poor productivity in 2009 continued to be exacerbated by the low success rate for drugs in late stage development and a decline in sales from new drugs launched within the last five years,” said Hans Poulsen, head of consulting at CMR International. Megan McArdle, the economics editor at The Atlantic, proffers two common explanations for the current dilemma, the first is that the FDA regulators are cracking down unreasonably on new filings. As an example, in July 2008, the U.S. Food and Drug Administration’s (FDA’s) Endocrinologic and Metabolic Drugs Advisory Committee voted in favour of a requirement to conduct a long-term trial or provide equivalent evidence to rule out an unacceptable cardiovascular risk for diabetes drugs that show no cardiovascular safety signals in phase 2 or phase 3 studies (1). The second is that the industry is reducing its R&D spend to improve the bottom line. Megan offers a third explanation that the “big fat targets in medicine have largely been hit” Fred Hassan (former CEO of Schering Plough) agrees stating that the next generation of targets are more complex and far more expensive to do R&D in.

So with productivity at a 10 year low, is the Big Pharma model broken, what is the future for the industry and is there a solution to problem(s)?

Historically, pharmaceutical companies have themselves led the way in medical research; knowledge generated through internal basic research programs or licensed from academic institutions to develop products that would in turn fuel the next generation of drugs. This leadership position was solidified by the substantial capital and infrastructure requirements of the drug development process that created significant barriers to entry for others. Academic institutions and small companies generally lacked the expertise, infrastructure, and financial resources to engage in drug discovery and development. To address slowing productivity, pharmaceutical companies initially sought the most rapid path to accessing new technologies and novel drug candidates, leading to a wave of industry mergers (see Figure 1).

These mergers failed to ignite productivity, which remains lower even as sales and R&D spending dollars have increased. Moreover, the number of new molecular entities (NMEs) approved today is down slightly over the same period (2). Only 19 new drugs or vaccines were approved in 2007, the lowest numbers since 1983.

A confluence of internal and external factors is now transforming the landscape for discovering, developing, commercialising, and marketing pharmaceuticals, and the old rules simply no longer apply to an industry now facing (1) pressure to increase sales, (2) pressure to decrease development time and cost, (3) competition from smaller companies, (4) looming patent expirations, (5) increased regulatory scrutiny, and (6) unparalleled pricing pressures. Companies now need to embrace two critical realities: a New Paradigm of Drug Development & Embracing Democratization: Partner or Perish.

The need for a new paradigm is self-evident, we have a bottleneck driven by the failure of many investigational therapies to demonstrate efficacy in phase 2 trials. These phase 2 failures are the result of an inherent flaw in the strategy of taking as many candidates forward as possible with minimal upfront investment in clinical validation. Rather than diversifying risk across multiple programs, this strategy has increased costs and the risk that an individual program will fail. If the pharmaceutical industry is to regain its prominence in advancing human health, companies need to take a leadership position in establishing a wholly new agile and pragmatic approach to delivering differentiated drugs to the market. Critically, this process must incorporate effective decision making to identify candidates with a high probability of success earlier in the development process to relieve the burden of late-stage failure.

Today’s experimental (or translational) medicine approaches generate insight into clinically relevant interactions among drugs, targets, pathways, and disease rather than reinforce existing hypotheses. Experimental medicine uses innovative, short-term clinical trials of small groups of human subjects to gain a preliminary assessment of pharmacological activity, efficacy, and safety of new chemical entities. Data from such trials should enable more rapid decision making earlier in the development process and make it possible to guide activities at the bench toward end points that truly matter at the bedside. Importantly, such trials enable early identification of populations more likely to benefit from a specific compound allowing stratification of downstream clinical trials in the hope of tailoring therapies effectively to improve the probability of success.

Merck has embraced the concept of failing quickly and cheaply because it shortens the timeline and substantially reduces the risk for those programs that do move forward. This approach has helped the company to reduce its phase 1 and phase 2 development timeline from the industry average of 3.7 years to 2.5 years (3). Our early-stage R&D process uses technology and outsourcing to generate as much data as quickly and cost effectively as possible. Elements of the company’s fail fast, fail cheap strategy include:

1.         Employing biomarkers and experimental medicine to look at changes in gene expression, protein activity, or metabolites that may be indicative of efficacy, off-target interactions, or toxicity. Although it may sound obvious that humans are not rodents, innumerable dollars have been spent pursuing programs that cure disease in animals but have limited benefit or demonstrate significant safety issues in humans.

2.         Incorporating more animal and human genetic studies early in the drug development process to better understand the role that potential targets play in the disease process and how they interact with other targets and pathways. These studies speak to target validation, can provide critical insight into interactions that might lead to unacceptable side effects or toxicities, and can help kill a program before it even enters the clinic.

3.         Conducting more extensive preclinical pharmacology and toxicology studies for much the same reason as noted above.

4.         Outsourcing medicinal chemistry capabilities to screen a greater number of candidates at a lower cost than could be achieved using internal resources.

In pursuing a philosophy of fail cheap, fail early, Merck was looking for technologies and processes that would allow to be able to make go/no go decisions as early in development as possible while stopping programs with little apparent probability of success. The company established the Experimental Medicine department alongside our Clinical Pharmacology group to create tools that enable good choices earlier in the R&D process. Both Experimental Medicine and Clinical Pharmacology can now use small, short-term clinical studies to provide early answers to important questions regarding pharmacological activity, efficacy, and safety of candidate drugs, including:

  • Target engagement: How hard and how long is the target being hit?
  • Proof-of-pharmacology: Does hitting the target elicit the intended biological effect?
  • Biomarker (linkage to clinical outcomes) qualification: Are the resulting changes meaningful in the context of the disease?
  • Disease Mechanism: How does the clinical data enhance the understanding of disease pathways and development and use of disease models?

In 2008, U.S. and European regulators announced that they will accept data from seven new biomarker tests designed to assess kidney damage in animal studies of investigational therapies. Merck scientists worked collaboratively with colleagues at other companies to identify the markers and demonstrate their utility in detecting renal toxicity within hours of exposure to investigational drugs. These biomarkers are a significant advance over the standard assays for renal toxicity, which detect kidney damage at least 1 week after it has started (4). The difference between hours and weeks can dramatically impact early-stage development timelines and enable rapid elimination of compounds likely to cause renal toxicity, a common factor in late stage drug candidate failure.

The cultural shift also facilitated the reallocation of financial and personnel resources from drug discovery to biomarkers and experimental medicine practices because the discovery teams recognised the value that these trials would bring to the portfolio as a whole. It also helped Merck’s scientists transition from a decision-making process predicated on gathering data on every aspect of a compound or target to one based on more limited (but more clinically relevant) data. Rather than aiming for 100% analysis, decision makers now follow the “80/20 rule” in determining if a program merits continued investment. The power of experimental medicine makes that 80% confidence level nearly as informative as 100% confidence based on less clinically relevant data. That 20% reduction can significantly decrease development time and costs because it enables the company to get into and out of phase 1 and phase 2 development more rapidly. The uncommonly rapid development and tremendous market uptake of Merck’s Januvia (sitagliptin) diabetes therapy demonstrates the benefits of this approach, particularly because the rapid timeline did not compromise the rigorous safety testing that is a critical component of all of Merck’s development programs.

The democratisation of drug discovery has lowered the barrier to initiating phase 1 trials, as evidenced by a two to threefold increase in the number of compounds in phase 1 trials and a similar increase in phase 2 trials (5). Biology has become “open source,” and pharmaceutical companies no longer control the resources and infrastructure for generating insight to disease and therapeutic strategies and advancing discoveries through the clinic. The emergence of a service industry that provides quality drug development resources outside of the pharmaceutical industry (including large compound libraries; state-of-the-art, high-throughput screening for both on and off target activities; standard absorption, distribution, metabolism, and elimination (ADME) testing capabilities; clinical research organisations; and pilot manufacturing facilities) has empowered individuals and institutions with innovative ideas, intellectual property, and financial backing to develop their own drug candidates rather than license technology to the pharmaceutical industry at an early stage. The ability to outsource most early-stage research functions has spurred an increase in the number of virtual companies, allowing inventors to remain engaged in advancing their own innovations at minimal cost. This democratisation simultaneously expands competition at the early stages of drug development while reducing the pool of intellectual property that big pharma can license for its own development initiatives.

Rather than posing a competitive threat, democratisation has created numerous and diverse opportunities for companies to tap into the entrepreneurial drive of these organisations while sharing both risk and reward. The industry’s ability to maintain its leadership in these areas rests on its ability to continue to convince earlier-stage companies and academic institutions that collaboration is a more compelling strategic option than pursuing commercial activities on their own.

Collaborations provide a biotechnology company with money and resources while providing the pharmaceutical company access to cutting-edge technologies. In addition, by collaborating with multiple partners, pharmaceutical companies today decentralise parts of their R&D activities. This decentralisation provides a mechanism by which companies can (1) evaluate multiple new platform or product opportunities without increasing the size and cost of their own operations and (2) effectively increase the bandwidth of their operations.

In 2010 64% of Merck’s revenue was attributable to alliance products and patents (includes 50% of all joint venture revenue). In the past 8 years, Merck has established in excess of 300 collaborations with leading academic centers and biotechnology companies. These collaborations have allowed the company to participate in breakthrough discoveries in important therapeutic areas, including Alzheimer’s disease, ophthalmology, cancer, bone disease, and psychiatric disorders.

Charles Darwin observed that it is not necessarily the strongest or most intelligent that survive but those most responsive to change. The pharmaceutical industry stands at an important juncture facing patent expirations, increased generic competition, increasing regulatory standards, and decreasing profit margins. Those companies that embrace change by seizing the opportunity for innovation are more likely to thrive and evolve over the long term. Fundamentally, the ability to adapt to a new environment is as much a matter of state of the art as it is state of mind for the pharmaceutical industry. At Merck, we have established four pillars that support an environment for change:

1.         Having a portfolio mindset that unifies all aspects of the organisation because it creates a shared mark of success. Motivating individuals and teams to focus on the success of the portfolio rather than on that of a particular compound greatly facilitates continued prioritisation and rationalization of the early-stage pipeline.

2.         Redefining failure. A failed trial or program is really a success if it happens early enough to allow advancement of better programs and provides insight that can improve future development efforts.

3.         Valuing quality science, regardless of its source.

4.         Recognising that owning part of a great product is ultimately more valuable than owning 100% of a failure.

These four pillars are applicable across diverse R&D models. With the advent of molecular biology as a critical component in the drug discovery and development process, pharmaceutical companies have generally pursued one of two divergent paths to commercial success. Some have opted to focus on specialty medicines for acute or serious diseases that have a clear molecular etiology, offer greater risk–benefit ratios, and have the potential to be used with a companion diagnostic test. Others continue to drive toward the challenge of the large, primary care markets in areas of major unmet medical need, such as metabolic disease and the neurosciences.

A key driver in choosing the path of specialisation is a belief in the power of personalised medicine. This belief derives from the simple but elegant logic of understanding the molecular basis for an individual’s disease and matching treatment to a specific disease mechanism. Nowhere is this philosophy more apparent than in oncology, in which targeted therapies addressing cancer-related signaling pathways are transforming both the understanding and the treatment of disease. Cancer is, by any definition, a genetic disease that results from an accumulation of mutations in affected tissue. Consequently, if personalised medicine is likely to succeed anywhere, it will likely be in oncology.

Despite the logic on which personalised medicine is founded, and the expanding base of knowledge focused on how specific genes and pathways contribute to the development and progression of disease and response to therapy, success with this approach in major, primary care diseases has been frustratingly elusive. Although a small number of examples give hope that personalised medicine will play a transformative role in these widespread diseases, the path to achieving that transformation remains unclear. For example, diabetes patients are still segmented by the temporal stage of their disease and degree of insulin resistance, even though these are symptoms of the disease rather than the underlying cause.

The industry-wide need for improved efficiency coupled with a growing demand among payers and physicians for demonstrable value will demand improved tools and strategies for matching patients with therapies. Although the industry continues to work hard to adapt to this new environment, a few early success stories have established new and better paths to success. Several products have provided a strong indication of how development times may be condensed. Using clinically relevant biomarkers early in development can dramatically shorten the timeline for late-stage trials, reduce the risk of clinical failures, and enable the development of new therapies. Early successes are at the leading edge of what must be a fundamental transformation.


1. Hughes S. FDA advisory committee recommends cardiovascular safety studies for diabetes drugs. The, July 3, 2008.

2. Global pipelines quantified. Merrill Lynch Research Report. August 23, 2007.

3. CMR International, 2001–2005 (Major Company Median Comparison). June 2006.

4. FDA, European Medicines Agency to consider additional test results when assessing new drug safety. FDA News, June 12, 2008.

5. Parexel International. (2007). Pharmaceutical R&D Statistical Source Book 2007/2008, edited by Mathieu MP. Waltham, MA: Parexel International.

About the Author

Phil Kearney is the Director of Licensing & External Research at MSD. After a long career in academia Phil moved to Scandinavia to take up senior and executive roles in firstly Swedish & then Danish biotechs, before returning to Australia in 2006 in his current position.